WORLD BANK TECHNICAL PAPER NO. 533 - - ,\ Europe and Central Asia Poverty Reduction and Economic -- , wManagement Series THE WORLD BANK WTP533 Work In progress I f or pubicdi,scussion January 2003 Expenditure Policies Toward EU Accession K. , . Edited by Bernard Funck Recent World Bank Technical Papers No 428 C Mark Blackden and Clutra Bhanu, Gender, Grozwth, and Poverty Reductiow- Special Prograni of Assistance for Africa, 1998 Statuis Report on Poverty in Sub-Saharan Africa No 429 Gary McMahon, Jose Luis Evia, Alberto Pasc6f-Font, and Jose Miguel Sanchez, An Environnmental Study of Artisanal, SiSiall, and Mediinni Mtining in Bolivia, Chile, and Peru No 430 Maria Dakou,,C, Coturt Performnance arouind the World A Comnparative Perspective No 431 Severin Kodderitzsch, Retfornis in Albanian Agricultutre Assessing a Sectoi in Transition No 432 Lutz Gabriel Azevedo, Musa Asad, and Larry D. Simpson, Managellment of Water Resources. Bulk Water Pricin7g in Brazil No 433 Malcolm Rowat and Jose Astigarraga, Latin Anmerican Inisolvenicy Systemls A Comzparative Assessnient No. 434 Csaba Csaki and John Nash, eds., Regional aiid International Trade Policy Lessons for the ELI Accession iii the Rural Sector-World Bank/FAO Workshop, Jiune 20-23, 1998 No 435 lam Begg, ELI Investnment Grants Review No 436 Roy Prosterman and Tim Hanstad, ed., Legal Inipedinuents to Effective Rural Land Relations in Easterii Eutrope and Central Asia A Conuparative Perspective No 437 Csaba Csaki, Michel Dabatisse, and Oskar Honisch, Food and Agricultutre in the Czech Repiblic Froln a 'Velvet" Transition to the Challenges of ELI Access iou No 438 George J Borjas, Econoniuc Research on1 the Determrinants of Imnuigration Lessons for the European Uniion No 439 Mustapha Nabli, Financial litegration, Vtilniabilities to Crisis, aiid EU Accessioi ini Five Central Europeaii Coiiii tries No 440 Robert Bruce, loannis Kessides, and Lothar Kneifel, Overconuumug Obstacles to Liberalizatiomi of the Telecomn Sector in Estoiiia, Polnrid, the Czech Repuiblic, Slovenia, amid Huni,gany An Overview of Key Policy Concernis atid Potential Initiatives to Faciltate the Trauisition Process No 441 Bartlomiej Kaminski, H-lungary Foreigui Trade lssiues in the Conitext of Accession to the EU No 442 Bartlomiej Kaminski, The Role of Foreigmi Direct Investnuient and Trade Policy in Poland's Accession to the Euiropean Uuiiomi No 443 Luc Lecuit, Jolhu Elder, Christian Hurtado, Fran,ois Rantrua, Kamal Siblini, and Maurizia Tovo, DeMlStlfyiiig MIS Guideliniesfor Managenment Informantitton Systenus in Social Fuinds No 444 Robert F Townsend, Agriciiltiial Incentives in Suib-Saharan Africa Policy Challenges No 445 Ian Hill, Forest Managenuent in Nepal Ecoiiouiiics of Ecology No 446 Gordon Hulighes and Magda Loveu, Econoniic Reform annd Envurouunuientuil Perforinlaiice tii Tramisitioni Ecotnioiies No 447 R Maria Saleth and Ariel Dinar, Evaluatitng Water Institiitions anid Water Sector Perform1ance No 449 Keith Oblitas and J Raymond Peter in association with Gautam Pingle, Halla M Qaddumi, and Jayantha Perera, Traiiisfe-rinilg Irrigatioii Manageniment to Farinieis iii Andhra Pradeslh, lidia No 450 Andres Rigo Sureda and Waleed Haider Malik, eds, Juidicial Challenges in the New Millentti int Proceediugs of the Second Sutintint of the Ibero-Anierucan Suiprenlue Couirts No 451 World Bank, Privatizationi of the Power and Natuiral Gas Iuidiistries tii Hiiiigary and Kazak/istali No 452 Lev Freinkman, Damel Treisman, and Stephen Titov, Subnattonal Buidgetiuig in Russia Preemnpting a Potential Crisis No 453 Bartlomle Kaminski and Ibchelle Riboud, Foreigni Investuiient and Restructuring The Evidence froimi Hi1iugary No 454 Gordon Hughes and Julia Bucknall, Poland Comlplying w7ith EU Environnmental Legislature No 455 Dale F Gray, Assessineiit of Corporate Sector Valiue amid Vullnerability. Links to Exchange Rate aiid Filiauicial Crises No 456 Salman M A Salman, ed , Groundwater Legil and Policy Perspectives Proceedings of a World Bank Seinunar No 457 Mary Canning, Peter Moock, and Timoth-y Helermak, Reforninlllg Education in the Regions of Ruissia No 458 John Gray, Kazakhstan A Review of Farni Restructutring No 459 Zvi Lerman and Csaba Csaki, Llkramne Reviewo of Fnirm Restructuring Experiences No. 460 Gloria La Cava and Rafaella Y. Nanetti, Albania Filliuug the Vulnerabuluiy Gap No 461 Ayse Kudat, Stan Peabody, and Caglar Keyder, eds., Social Assessm1ent aiid Agricultural Reforni in Central Asia anrd Turkey No 462 T Rand, J Haukohl, and U Marxen, MunIicipal Solid Waste Incinerationi Requirenuenits for a Successfiul Project (List continues on the inside back cover) WORLD BANK TECHNICAL PAPER NO. 533 Europe and CentralAsia Poverty Reduction and Economic Management Series Expenditure Policies Toward EU Accession Edited by Bernard Funck The World Bank Washington, D.C. © 2002 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved. 1234050403 Technical Papers are published to communicate the results of the Bank's work to the development com- munity with the least possible delay. The typescript of this paper therefore has not been prepared in accor- dance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed here are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank cannot guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply on the part of the World Bank any judgment of the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. No part of tiis work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or inclusion in any information storage and retrieval system, without the prior written permission of the World Bank. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint, please send a request with complete information to the Copy- right Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. ISBN: 0-8213-5368-3 ISSN: 0253-7494 Bernard Funck is Lead Economist in the Europe and Central Asia Region at the World Bank. Library of Congress Cataloging-in-Publication Data has been requested Contents Foreword .............................................. vi Abstract .............................................. vii Acknowledgments .............................................. viii Executive Summary .............................................. ix 1. Expenditure Trends and Challenges ...............................................1 Fiscal Objectives of the CEECs ..............................................1 Recent Trends in General Government Expenditure ...............................................3 Differences in Expenditure Composition ...............................................7 Conclusions .............................................. 14 2. Liquidating the Legacy of the Past .............................................. 15 Restructuring Banks .............................................. 15 Restructuring Enterprises .............................................. 22 Conclusions .............................................. 30 3. Successfully Integrating with the European Union ......................................... 31 Upgrading Skills .............................................. 31 Upgrading Environmental Standards .............................................. 49 Upgrading Transport Networks .............................................. 58 Conclusions .............................................. 68 4. Ensuring Social Protection .............................................. 71 Reforming Public Pension Systems .............................................. 71 Reforming Health Care .............................................. 84 5. An Expenditure Strategy for Growth and Convergence .............................................. 99 List of Tables Table 1.1 Pre-Accession Fiscal Programs of the CEECs, 2000-04 .2 Table 1.2 Fiscal Cost of EU Accession and Sources of Financing: National Estimates . 2 Table 1.3 Consolidated General Government Expenditure (Excluding Net Lending), 1995-2000 .3 Table 1.4 Central Government Debt and Interest Payments .4 Table 1.5 Consolidated General Government Primary Expenditure, 1995-2000 .5 Table 1.6 Consolidated General Government Expenditure (Excluding Net Lending) by Economic Type, 1995-2000 .8 Table 1.7 General Government Expenditure (Excluding Net Lending) by Function, 1995-2000 .9 Table 1.8 Government Employment and Expenditure on Wages and Salaries, 1996-2000 .11 Table 1.9 Growth of the Age 0-14 Population. 11 Table 1.10 Poverty Incidence and Welfare Payments, 1995-2000 .14 Table 2.1 Shares of Nonperforming Loans .15 Table 2.2 Total Fiscal Costs of Bank Restructuring in Selected CEECs, 1991-98 .17 iii Table 2.3 Assets of Majority Foreign-Owned Banks .................................................................. 18 Table 2.4 Banking Sector Reform and Financial Regulations .................................................................. 21 Table 2.5 Key Elements of the Banking Institutional Framework ............................................................. 22 Table 2.6 Enterprise Priva tizatio n and Restructuring, 2001 .................................................................. 24 Table 2.7 Rail Transport Volumes and Labor Productivity, 1999 ........................................................... 26 Table 2.8 Operating Deficits of Selected CEEC Railways, 1997 ................................................................. 27 Table 3.1 School-Age Population, 1990-2015 .................................................................. 32 Table 3.2 Enrollment Ratios through the Transition .................................................................. 33 Table 3.3 Bulgaria: Rates of School Attendance by Level, 1995, 1997, and 2001 ..................................... 34 Table 3.4 Public Expenditures on Education, 1990-2000 .................................................................. 35 Table 3.5 Changes in Student/Teacher Ratios .................................................................. 36 Table 3.6 TIMSS Eighth Grade Student Assessment Results for Science and Math for Eight CEECs, 1995 and 1999 ............................................................... 37 Table 3.7 PISA Student Assessment Results for Literacy for 15-Year Olds for Four CEECs, 2000 ....... 38 Table 3.8 A Composite Formula for Education Finance ............................................................... 41 Table 3.9 Total Benefits over the Period 2005-20 ............................................................... 49 Table 3.10 Estimated Costs of Environmental Acquis for the CEEC ............................................................ 50 Table 3.11 Ratio of Benefits to Costs of Complying with Environmental Directives ................................ 52 Table 3.12 Alternative Estimates of Investment Costs for Compliance ..................................................... 53 Table 3.13 Replacement Value of Main Road Network, 1999 ............................................................... 59 Table 3.14 Internal Rate of Return oun Road Projects: Best Practices ........................................................... 60 Table 3.15 Tolls and Vignettes in the CEECs ............................................................... 63 Table 3.16 Earmarking of Road User Revenues ("Road Fund") in the CEECs .......................................... 65 Table 3.17 Road Fund SWOT ............................................................... 67 Table 4.1 Deficit/Surplus of Public Pension Systems in the CEEC, 1991-2000 ........................................ 72 Table 4.2 Expenditures of Public Pension Expenditures in the CEECs, 1991-2000 ................................. 72 Table 4.3 Basic Characteristics of Mandatory Pension Systems in CEEC Countries Following Transition Periods .76 Table 4.4 Projected Financial Performance of the Public Pillar following Reforms approved through End-2001 .................................................................. 80 Table 4.5 Public Pension Expenditures in EU Member Countries: 2000-50 ........................................... 84 Table 4.6 Trends in Health Care Expenditures in the CEECs, 1990-99 ..................................................... 86 Table 4.7 Comparison of Health Status Indicators between the CEECs and the EU .............................. 87 Table 4.8 Share of Social Health Insurance and Taxation in Public Spending on Health ....................... 88 Table 4.9 Social Insurance Contribution Rate for Health .................................................................. 90 Table 4.10 Paying Health Care Providers .................................................................. 91 Table 4.11 Hospital Beds, Physician, and U tiliz ation Patterns .................................................................. 92 List of Figures Figure 1.1 General Government Expenditure and Income per Capita in the EU and the CEEC, 1999 ....6 Figure 1.2 Consolidated General Government Outlays in the CEECs and the OECD 1995-2000 ............7 Figure 1.3 Government Wage Bill and Perception of Corruption ............................................................... 12 Figure 3.1 Condition of Road Network, 1999 .................................................................. 59 Figure 3.2 Average Daily Traffic Density on Motorways and Two-Lane Major Roads ........................... 60 Figure 4.1 Demographic and Pension System Dependency Ratio .............................................................. 73 Figure 4.2 Number of Beneficiaries from Public Pension Systems, 1991-2000 ......................................... 74 Figure 4.3 Number of Contributors to the Public Pension System, 1991-2000 .......................................... 75 Figure 4.4 Demographic Dependency Ratio, 1995-2040 (population above 60/Population 20-60) ........ 77 Figure 4.5 Health Care Expenditure in the CEECs, 1990-99 .................................................................. 85 iv List of Boxes Box 2.1 Restructuring the Polish Coal Mines ......................................... 25 Box 2.2 Restructuring Polish Railways ......................................... 28 Box 2.3 Restructuring the Energy Sector in Romania ......................................... 29 Box 3.1 Bank Experience with Water Utilities in the Baltic States ......................................... 54 Box 3.2 Hungary Ml / M15 Motorway Concession Project ......................................... 62 Box 3.3 Road User Charges in the CEECs .......................................... 4 Box 3.4 Key Features of the Latvian Road Fund ......................................... 66 Box 4.1 Indebtedness in the Health Systems ......................... ..................................... 96 v FOREWORD The Poverty Reduction and Economic Management Unit in the World Bank's Europe and Central Asia Region has been undertaking a series of analytical works on the issues pertinent to the economies in the region. These issues indude transition issues, issues of economic integration pertinent for the Central and Eastern European countries that are candidates for accession to the European Union; poverty issues, decentralization issues, and other economic management issues. The analytical work has been conducted by staff of the unit and other bank staff, as well as specialists outside of the bank. This technical paper series was launched to promote wider dissemination of this analytical work, with the objective of generating further discussions of the issues. The studies published in the series should therefore be viewed as work in the progress. The findings, interpretations, and conclusions are the authors' own and should not be attributed to the World Bank, its Executive Board of Directors, or any of its member countries. Cheryl Gray Sector Director Poverty Reduction and Economic Management Unit Europe and Central Asia Region vi ABSTRACT This report forms part of a larger project that is entitled, "EU Accession-Developing Fiscal Policy Frameworks for Sustainable Growth," the theme of a conference organized by the European Commission (and cosponsored by the World Bank and the International Monetary Fund) in Brussels on May 13-14, 2002. The purpose of this report is to discuss the set of public expenditure policies that might be conducive to rapid growth and convergence among the Central and Eastern European countries. It was left to others- complementary contributions to discuss two other key dimensions of expenditure reforms: the overall macroeconomic framework in which they take place and to which they contribute and the institutional and political economy conditions under which successful reform strategies can be designed, find political support, and be implemented. In this report, we seek to take stock of the countries' own public expenditure policy objectives and to distill the best practices and lessons learned in designing of expenditure reforms within those countries. It is obviously impossible within the limits of a report such as this one to address every expenditure program. The option was thus taken to focus on some of the largest or most sensitive expenditure items, as well as those most affected by the European Union accession process. We conclude that the general thrust of the expenditure strategies candidate countries have put forward in their pre-accession economic programs appears both appropriate and at least theoretically feasible. A primary tool for curbing expenditure should be to rein in subsidies and transfers. The report also highlights ways in which key expenditure programs (such as on education, transport and the environment) could be redirected to be more fully supportive of growth objectives. Factors related to a country's political economy and to the institutional framework of public resource management will undoubtedly play a determining role in framing what actual policy choices will eventually be made. 1.1 vii ACKNOWLEDGMENTS This study has been prepared by a team led by Bernard Funck. The team comprised Mukesh Chawla, Elena Kastlerova, Esperanza Lasagabaster, Anil Markandya, Michael Mertaugh, Eva Molnar, Helena Tang, Giedre Taburniene, and Louis Thompson. The team also received useful advice, input, and comments from Jean Jacques Dethier, Armin Fidler, Roger Grawe, Cheryl Gray, Franz Kaps, Maureen Lewis, Catalin Pauna, Kyle Peters, and Michal Rutkowski. The report was processed by Anita Correa. viii EXECUTIVE SUMMARY In their pre-accession economic programs (PEPs), most Central and Eastern European countries (CEECs) have put forward medium-term fiscal strategies to improve budget balances and cut back taxation levels. The corollary is a need to bring down public expenditure by an average of 2.5 percent of gross domestic product (GDP) over the period 2000-2004. The National Programs for the Adoption of the Acquis (NPAA), on the other hand, consider that accession may entail additional expenditures to the tune of 3.5 percent of GDP on average. These two conflicting sets of objectives can be reconciled only through deep cuts in regular expenditure prograrns. Unfortunately, there is little evidence that would demonstrate a common resolve to bring this about: on average, the levels of primary expenditures (that is, net of interest payments) remained essentially unchanged as a percent of GDP between 1995 and 2000. With a few exceptions, this general observation is also valid at the individual country level. Are these objectives therefore out of reach? Perhaps not. First, there is evidence that at least some countries have managed to curb their expenditure levels in the recent past. Second, as this report will point out, there is sufficient variation across the expenditures of CEECs by types of expenditure to underpin the notion that expenditure reform is at least theoretically feasible, even though it is politically difficult. The purpose of this report is to take stock of that experience and of the measure of upcoming challenges and then to distill the lessons that emerge from the early efforts of the CEECs to reform their expenditure policies. Examples of best practices will emerge along the way. We will review recent trends in comparative aggregate expenditure levels, and then shift to a more disaggregated approach to search out the flexibility to affect these trends. On this basis, the subsequent parts of this paper examine the ways in which CEECs are dealing with the three major expenditure challenges confronting them: * Liquidating the legacy of transition (that is, recognizing, containin& and financing the costs of bank and enterprise restructuring). * Helping economies rise to the challenges of European integration by upgrading skills levels, environmental standards, and transport networks. * Providing social protection to an aging population in an affordable manner, particularly in the two key areas of pensions and health. ix Expenditure Levels and Trends While CEECs have often made some progress in bringing down total government outlays, the record of curbing primary expenditure is at best mixed. The only major examples are those of Estonia (which compressed spending on goods and services) and Hungary and the Slovak Republic (which managed to cut back current transfers). On average, however, primary expenditure remained steady at about 38 percent of GDP over the period. This level is about the same level as that of the cohesion countries (Greece, Ireland, Portugal, and Spain). A closer look at the data reveals, however, (a) that the traditional positive statistical correlation between income and government size is actually inverted in the case of the cohesion countries and (b) that the country in which government is the smallest-Ireland-is also the country in which income per capita has risen the most rapidly since the mid-1980s. To the extent that a causal link exists in these relationships, it is Ireland's experience that the CEECs should naturally be tempted to replicate. Indeed, the general trend of government expenditure across the Organization for Economic Cooperation and Development (OECD) has been downward: The trend has been most pronounced within the European Union (EU) where general government outlays shrank by five percentage points of GDP between 1995 and 2000. The relative decline was perhaps less outside of Continental Europe, but the starting point was also considerably lower, bringing into even sharper contrast the relative performance of the CEECs. By 2000, general government outlays in the CEECs exceeded the levels observed among non-Continental European members of the OECD by a full seven percentage points of GDP. The variation in expenditure levels across CEECs lends support to the view that, far from being pre-ordained, expenditure levels are subject to a fair degree of policy discretion. In 2000, primary expenditure varied between around 43 percent of GDP in the Czech Republic and Slovenia to slightly more than 30 percent in Lithuania and Romania. This observation stands even after controlling for differences in income level. An even greater variation can be observed when government expenditure is broken down into economic and functional categories. For just about any particular item of the Governance Finance Statistics (GFS) classification, the levels of expenditure (relative to GDP) vary by at least a factor of two, and the differences range all the way up to a factor of twelve in the case of enterprise subsidies. Such variations seem to provide evidence that scope exists to carry out the type of expenditure reform implicit in the CEECs' Pre-Accession Economic Programs. A closer inspection of the expenditure categories gives a first set of clues as to where the main opportunities for reform might be. Such a closer look shows that the CEECs: r Spend less than member countries on wages and salaries; their social sectors, however, are often overstaffed. o Are already catching up with (and sometime exceeding) the high levels of public investment observed amnong cohesion countries, which raises concerns about the quality of the project pipeline and the possible displacement of operations and maintenance expenditure. x * Devote the lion's share of their public resources to funding transfers and subsidies. Early reformers are beginning to see the benefits of their efforts, but pressures are mounting on the later reformers. * In some cases, have not yet managed to rein in subsidies for enterprises (broadly speaking), in part because of the ongoing cost of bank restructuring. How to bring reform about in practice is a different matter. To answer that question, we turn to the real-life experience of the CEECs in dealing with the three key expenditure challenges mentioned previously, bank and enterprise restructuring; upgrading of skills, environmental standards, and transport networks; and provision of social protection. Liquidating the Legacy of the Past The inability of most CEECs to enforce financial discipline on enterprises at the beginning of transition has left - a painful and burdensome legacy. Banking problems constituted only one manifestation of the underlying enterprise problems. In addition to "soft" loans from the banking sector, unviable enterprises were kept afloat through arrears to the state (most prominently tax arrears and arrears in social security contributions) as well as arrears to other enterprises (most prominently public utilities). For many CEECs, liquidating this legacy remains at the top of the list of fiscal priorities. Banking Sector Restructuring A key issue facing the CEEC governnents is how to minimize fiscal costs while liquidating their liabilities. These costs can be very high. For the period 1991-98, the total fiscal costs of such crises for some of the CEECs ranged from 2 percent of GDP for Estonia to 42 percent for Bulgaria. The bulk of the fiscal costs have been direct costs incurred by the government to recapitalize banks. The following broad lessons emerge: * The earlier that portfolio problems are openly recognized and addressed the better. One main reason why the Slovak and Czech bailout operations were so much costlier than the Estonia bailout is that in the Slovak Republic and the Czech Republic the underlying bad loan problems were left to fester for a long time before being addressed frontally. * Successful bank restructuring ultimately requires bank privatization, preferably to strategic investors. Foreign banks in particular have played a central role in breaking up the "old boys' club" that used to bind banks to their borrowers and that proved fatal to the banks. The now massive presence of foreign banks (they control more than two-thirds of bank credit in most CEECs) should help the banking industry grow on a sounder basis. * Banks need to be fully recapitalized first for bank privatization to be successful. Otherwise the risk is high that the banking regulator will show excessive regulatory forbearance, or that the new private owners will see no better option than to engage in the same type of reckless behavior as their predecessors, sowing xi the seeds of continued banking problems (as the examples of Hungary, Bulgaria, and the Czech Republic have shown). The generally healthier-looking balance sheets of commercial banks today provide a clear sign of how much progress has been made. This does not mean that the bad loan problem has vanished. In many countries, the reason the numbers look smaller is that nonperforming portfolios have been carved out and transferred to a specialized debt recovery agency. Thus, how to dispose of these bad loans remains a key issue. Two approaches (one centralized and the other decentralized) have been tried, without clear evidence that one is superior to the other in terms of recovery rate or enterprise restructuring. In the circumstances, an eclectic, multi-track approach may actually bring about the best results. Hungary, for instance, was quite successful with its mixed private sector/state strategy involving both centralized and decentralized approaches. The Slovak Republic has recently adopted a similar multi-track approach involving sales of pools of loans to private investors, auctioning of individual loans to smaller investors, outsourcing of collection through asset/legal management contracts, and selective writing off of noncollectible claims. No matter which approach is adopted, effective creditor rights are essential, in terms of recoveries under bankruptcy and also in terms of incentives for both debtors and creditors to settle to avoid bankruptcy. Unfortunately, building a full-blown bankruptcy system has often taken a painfully long time. For this reason, the Czech Republic and the Slovak Republic are both focusing first on improving liquidation, promoting out-of-court restructuring mechanisms, and introducing a fast restructuring or reorganization track. Enterprise Restructuring Restructuring enterprises is the other side of the coin. One important route through which budget constraints have been hardened is privatization. While almost all of the CEECs have reached industrialized market economy standards in terms of small-scale privatization (with only Bulgaria and Romania lagging slightly behind), progress has been slower for larger firms: by 2001, fewer than 50 percent of the large-scale enterprises had been privatized in five out of ten CEECs. In cases involving large externalities, governments may need to incur fiscal costs in the short term to deal with the social consequences of restructuring large-scale enterprises in order to reap the longer-run benefits that privatization would ultimately bring. The restructuring of coal mines provides an example of what such a task entails: The cost of the (to date largely successful) Polish program (1998-2002) is amounting to around US$2 billion (equivalent to 1.3 percent of GDP), mainly for social measures. Railways are another case in point. In most CEECs, railways have suffered from a decline in the demand for rail transport (linked to the shift from transport-intensive heavy industries toward a more service-based economy and the emergence of more competitive modes of transport-trucks for freights and cars and airlines for passengers-without a corresponding reduction in staffing). Massive restructuring and labor shedding will often be required to halt the resulting drain on public budgets. Such large-scale restructuring (also xii mandated by the adoption of the acquis communautaire) will be costly, particularly for labor redundancy reduction. Restructuring the entire Polish railway system, for example, may eventually require a 50 percent reduction in employment, totaling about 300,000 jobs, and US$1.5 billion (0.9 percent of GDP). This is a matter of mere survival, however, and, as in the case of coal, footing the bill early will bring about much larger fiscal benefits later. A final major factor in softening the budget constraints on enterprises in some CEECs is arrears to other enterprises, in particular to state-owned public utility companies. Here, again, the adoption of the acquis communautaire is helping, because it encourages such measures as the unbundling of utilities, the restructuring of tariffs and the elimination of cross-subsidies, and open access to networks. Lithuania's power sector, for example, completed this process in January 2002. Going beyond the acquis communautaire, however, many CEECs are now actively mobilizing the private sector to participate in the provision of infrastructure services. Telecommunications, gas, and power utilities have been or will soon be privatized in countries such as Lithuania, the Czech Republic, Hungary, and the Slovak Republic. It is too early to assess the results of these recent privatizations. If the experience of other countries around the world is any guide, then private sector participation in the provision of infrastructure services, if well developed, can boost efficiency and ease the strain on public finances, while increasing the coverage of such services. Successfully Integrating with the European Union As the burden of the legacy of the early transition period is reduced, governments around the region are facing new expenditure demands-generally legitimate but often overwhelming-to support their countries' integration into the EU, and beyond, into the global economy. Many of these demands relate to the need to upgrade skills, environmental standards, and transport networks. The nature of the fiscal challenges involved differs from one case to another. There is generally a clear need for countries to invest in environmaent and transport. Some countries are already investing heavily. Governments will need to be all the more vigilant in that some of these demands come supported by financing, sometimes on very attractive terms. Even in those cases the issue is still to ensure that higher priority claims on public resources are not displaced and that the programs being financed are sustainable beyond the initial investment. In the case of education, however, the issue is the more difficult one of redirecting expenditures from yesterday's priorities to those of today and tomorrow. That is, while the benefits of expanding parts of the education system may be evident, shrinking other parts of the system might be politically crippling. Upgrading Skills Human capital has been and will remain a key asset for securing the benefits of European and global integration. Together with supportive macroeconomic and financial policies and infrastructure investments, education will continue to play a key role in supporting xiii the transformation of the CEECs from low-income, resourced-based economies to high-income, knowledge-based economies. In the last decade, the demand for education was affected by two major changes: * There was a sea change in the demand for skills that rendered the skills of many manufacturing and mining workers irrelevant to the new needs and created a demand for other skills, particularly in the service sector. Furthermore, the decompression of wages that accompanied the transition boosted the implied returns on education, hence the demand for education. * There was a contraction in the school age population at rates formerly seen only in cases of war, famine, or pestilence. By the year 2015, the school age cohort in the CEECs will be from one-third to one-half smaller than it was in 1990, and about 25 percent smaller than it was in 2000. The result has been that enrollment rates rose at all levels except secondary education, while enrollment numbers below tertiary levels shrank markedly. These developments call for vocational secondary education to teach more generic skills, higher education to expand and become more flexible, and the legal and fiscal environment to encourage employers and local governments to develop lifelong learning programs to meet local (and global) skill needs. The needed redirection of education systems took place in a context of limited resources. The levels of public expenditure on education have remained broadly stable (relative to GDP) in 1995-2000. Of course, governments typically sought to secure additional resources for education, and sought to diversify financing sources through such actions as: (a) decentralizing education responsibilities to regional and local governments, (b) requiring parents to purchase textbooks and other educational materials, and (c) expanding private education. But the resources available to the sector generally remained limited. Under the circumstances, the success of the strategy for redirecting education just outlined was predicated on the capacity of governments to redirect resources from lower to higher levels of education, in part by downsizing staff and facilities in many primary and secondary education programs while reorienting them, improving their quality, and improving the salaries of the remaining teachers. The supply side, however, was slower to react to these new conditions. Indeed, the response of education managers throughout the CEECs has more often than not focused on maintaining the jobs of teachers and other educational staff. Student/teacher ratios (already below OECD levels at the start of the decade) fell further, and teacher workloads dropped to well below OECD standards (for example, an average 583 hours per year in primary school in Hungary compared to 788 hours in the OECD as a whole). One of two outcomes (or a combination of both) typically resulted in: (a) falling teacher salaries leading to a serious problem with teacher morale and motivation (as well as to growing problems of corruption in some of the CEECs); and/or (b) outlays for salaries and benefits rising as a share of other recurrent education expenditures and displacing expenditures vital to maintain and update teaching, learning materials and school infrastructure. xiv The feeble incentive for efficiency conveyed by the prevailing input-based financing formula and, in many countries, by the small size of the local jurisdictions in charge of schools, have further hindered the necessary restructuring. Partly as a result of these problems, both quality and access have begun to show signs of eroding, although from levels that are still relatively high. With a few notable exceptions (including Hungary and Latvia), student scores on the international TIMSS tests have deteriorated. In addition, PISA tests indicate a growing variation in quality between schools, and perhaps more worrisome, a relatively poor mastery by CEEC students of the higher-order skills needed for most of the fastest-growing jobs in the global economy. Furthermore, while the regional averages are encouraging, the detailed data reveal consistently lower quality and attendance in rural areas, among ethnic minorities, and among children from poor households. The data also show declining secondary education coverage in five countries (Bulgaria, the Czech Republic, Latvia, Lithuania, and, especially, Romania). Most of this decline has occurred in vocational and technical education. The decline in secondary enrollments reflects the perception that vocational education no longer ensures employment for graduates. It may also reflect pressures for some students to enter the labor market in order to augment falling household incomes and, in the poorest of the CEECs, the impact of the shifting of financing responsibilities to households for textbooks and other educational inputs, or of prohibitive "informal" payments. In response to these quality and equity concerns, new policy patterns are emerging in some countries that emphasize * Uniform nationwide education financing based on capitation formulas. Although they are difficult to calibrate, composite formulas that reflect cost differences among different programs, place-specific cost factors, or cost differences arising from special learning needs of students seem to be the way of the future. * Assignment of education matters to larger jurisdictions, or inducements for smaller jurisdictions to cooperate in the direction of economies of scale. The disappointing results on standardized international aptitude tests have also called attention to the need to upgrade education systems to the requirements of the global knowledge economy. National governments will need to * Exercise greater leadership in bringing about curriculum reform (as in Slovenia). * Consolidate vocational education programs into fewer specializations for clusters of kindred occupations (Hungary provides a good example). * Define appropriate mechanisms to finance the quality higher education that is needed to support the CEECs' move to innovation-driven growth. This may involve concentrating state budget financing on either the most capable institutions or the most capable students, and letting (duly accredited) private institutions and tuition payments look after the rest of the sector. xv Upgrading Environmental Standards In seeking to join the EU, the CEECs are also binding themselves to much higher environmental standards. Meeting the environmental acquis will bring considerable improvement to the quality of life. According to recent studies, the expected benefits from meeting air standards are generally the highest, followed by the benefits arising from stricter water standards. These benefits will unfortunately come at a cost. Indeed, discussions on the environmental acquis often revolve around the cost of meeting the environmental requirements for accession. The initial calculations made by the European Commission in 1997 suggested that the CEECs would have to devote on average about 5 percent of GDP to environmental- related expenditures (investment and operations and maintenance). This would be more than double the levels of such expenditures in the CEECs in recent years. While benefits are generally expected to exceed costs, this is not always the case. In particular, the benefits of the waste directive are generally below costs. Each individual investment will therefore need to be subjected to thorough cost/benefit analysis. While the acquis communautaire is prescriptive on environmental standards, it leaves considerable latitude on how to meet these standards. The price of complying varies accordingly. For instance, more detailed or recent studies, such as the ISPA (Structural Policy Instrument for Pre-Accession) strategy for Lithuania, come up with cost figures that are as much as 75 percent lower than the earlier calculations. For Bulgaria, the World Bank estimates costs at about half those of an earlier European Commission study. A number of measures can indeed be taken to improve the cost-benefit tradeoffs, induding adopting least-cost investment solutions (especially for energy-related investments), opening up procurement to international tender, and phasing in compliance measures over time, starting with those that willbring the most immediate benefits to the populations and scheduling the more expensive items later in time. Funding environmental investments, however, is not merely a matter of increasing expenditure commitments. Rather, there is a need for administrative changes in budget implementation to make funding more efficient (for example, consolidation of small ecological funds). Commercialization could potentially be helpful but still entails some budgetary burden. A more effective means of furthering EU compliance is likely to be through privatization of some of the larger and more polluting industries in CEEC countries. A clear understanding of the environmental liabilities of past damages is necessary, and a legal agreement with the government explaining the new owner's responsibilities is crucial to attracting credible international investors. Upgrading Transport Networks To spur growth, the CEECs are also called upon to upgrade and redirect their transport networks to the requirements of European integration. This report will focus on roads and xvi motorways because combined they take the lion's share of the investment programs. Key challenges include: * A rapid increase in motorization and a drastic shift in mode. Over the past decade the demand for road transport has grown radically, owing to the sharp decline of rail transport and the unprecedented growth of passenger car ownership. The number of vehicles in the CEECs increased on average between 70 and 120 percent in recent years. The supply side, however, has not kept pace. The road sector has been a victim of underinvestment in maintenance and modernization. As a result, the roads are mostly congested, slow, polluted, and unsafe. * European integration challenges. The most important EU accession challenges arise from the liberalization of international trucking, the increase in axle loads to the higher EU standards (11.5 ton/axle), and the accelerated corridor development programs under the Trans-European Network program. Although in most countries the roads needed to stimulate domestic economic growth are largely the same as those needed for international transit traffic, the attempts to expand the motorway network too rapidly and to standards that are not always economically justified have already diverted funds from maintenance and development. Indeed, even in the main corridors the rapidly increasing volume of heavy truck transport has not been matched by a corresponding expansion of capacity or strengthening of the pavement. While it is true that potential investors might be deterred by the current poor road accessibility, the CEECs increasingly recognize that the motorway construction program must be sustainable. When the construction and financing of the motorway section is planned its socioeconomic impact should be compared to investing in the equivalent amount in the preservation and improvement of existing road assets or investing in transport services or education or health. Road investments-whether to preserve the existing road assets (deferred maintenance and routine and periodic maintenance) or to add capacity to the basic network or to expand the motorway network-need to compete for limited funds. At present, few motorway sections have sufficient level of traffic to justify the investments. In contrast, returns tend to be very high on maintenance expenditure, followed by rehabilitation and pavement strengthening. It is therefore the rehabilitation and pavement strengthening that ought to be prioritized, if countries are serious about growth and convergence. Ensuring Social Protection Parallel with the broad consensus that exists in favor of market economics and integration with Europe and the rest of the world, there remains a widespread demand for governments to protect people against the major risks of life. Health and old age are two particularly sensitive areas. Since the early days of transition, however, governments have had to contend with the rising costs of extending such social protection. In a bid to curb these costs, one country after another has revisited the old Bismarckian model of social policy and has tried to define a new role and new models for the concept of social insurance that lies at its heart. xvii Reformning Public Pension Systemns Pension systems came under serious strains during the 1990s, and, despite various attempts at reform, pension schemes were still displaying deficits in most CEEC countries at the end of the decade. The causes of these deficits varied between essentially two groups of countries. Pressures emerged either on the expenditure side or the revenue side (or both), as beneficiaries expanded and/or the contribution base shrank. In most countries (particularly Romania, Lithuania, and Poland), the expansion in the number of pensioners exceeded the underlying growth in the elderly population, as early retirement, occupational privileges, and disability claims surged. Some countries suffered an even more damaging erosion of their tax base, particularly in the early transition years, as the labor force participation shrank and unemployment soared. While a few countries (for example, Poland, the Czech Republic, and Slovenia) were able to maintain a comparatively stable contribution base, and while Hungary, Latvia, and Lithuania largely stabilized their contribution base by the end of the decade, the contribution base has continued to fall sharply in Bulgaria and Romania through the second half of the decade. These two countries had increased contribution rates to cope with financial deficits and further rising labor costs and increasing incentives to remain in the informal market. Reflecting these pressures, system dependency ratios (the ratio of pensioners to contributors) fell throughout the region. The degree at which system dependency ratios deteriorated varied across countries, rising all the way up to 100 percent in Romania and Bulgaria. Accordingly, the CEECs fall into two broad clusters: * A first cluster of countries (including Bulgaria, Estonia, Latvia, Hungary, and Poland), where the resulting financial pressure was more keenly felt, undertook systemic reforms. In reforming their public pillar, they have basically pursued one of the following two approaches. In some countries, while the structure of benefits was changed, the defined benefit principle was preserved. In a second subgroup of countries (Latvia and Poland), the public pillar was transformed into a notional defined contribution scheme, a new pay as you go paradigm developed in Europe in the 1990s. Beyond reforming their public pillars, a number of CEECs also proceeded to establish second, funded, pillars, the size of which (as well the switching strategy) has generally been defined in consideration of the "transition costs" that reforms would generate. * Another duster (including the Czech Republic, the Slovak Republic, and Slovenia), where pressure to reform was initially less intense since system dependency ratios did not initially deteriorate as severely, limited their actions to more incremental parametric changes. xviii The outcome of these diverging policy courses is already quite visible in the aggregate data. In 1995, the CEECs were essentially polarized in two groups with a group of high spenders (Hungary, Poland, and Slovenia) devoting 18-20 percent to social welfare and the rest that were spending around 10 percent of GDP. By 2000, the picture had changed with two groups of countries converging toward the middle of the range as early reformers (Hungary, Poland) were set on a downward expenditure trajectory while spending was drifting upward among some of the former low spenders (the Czech Republic, Bulgaria, and, to a lesser extent, Estonia and Lithuania). Furthermore, actuarial projections presented in this report show that the pension schemes in countries belonging to the first cluster (Bulgaria, Estonia, Hungary, and Latvia) are better prepared, as a result of these reforms, in confronting the aging demographics (even though some additional reforms may still be necessary in attaining a balanced position in the long term). In contrast, while limited parametric changes at first succeeded in keeping the pension finances of countries within the second cluster within bounds, there is now a growing realization that, left unattended, their pension system deficits are set to expand at a rapid pace. The future policy agenda varies accordingly between those countries facing the challenge of consolidating recent reforms and those that still need to define a feasible approach to reform. The first cluster (Bulgaria, Estonia, Latvia, Hungary, and Poland) will have to resolve three key issues: enhancing coverage and compliance, consolidating the reformed public pillar, and promoting the safe development of private pension plans. Countries in the second cluster (including Lithuania, the Czech Republic, the Slovak Republic, and Slovenia) are now grappling with reform options. The new wave of reforms should seek to restore long-term sustainability, and not only to contain deficits temporarily. To achieve this, some form of systemic action of the sort adopted in the first group will be required, because there is a risk that continued incremental parametric measures would eventually erode trust in the public pillar. Those who are turning their attention to reform have a rich body of experience to draw upon. They can also draw strength from the fact that early reformers are now emerging with systems that are better placed to face the challenge of the aging demographics, and are on a stronger long-term financial footing than pension schemes in many EU countries. Refortning Health Care Following a decade of escalation in the cost of health care, almost all the CEECs have by now transformed the basic setup of their health care system in a bid to raise additional resources for the sector and to curb costs through efficiency gains. These objectives would be achieved through the separation of financing from provision of health care, the introduction of social health insurance, the establishment of new provider payment mechanisms, decentralization, and competition among providers and greater choice for patients. In many countries the reforms are too recent to assess. A few tentative observations nevertheless emerge. While countries have been experimenting with various provider payment arrangements, there is little sign as yet that health costs have been successfully contained. xix Measured in dollar terms (at purchasing power parity), health expenditure per capita surged by more than 50 percent during the past decade. This compares to a 25 percent increase on average among the cohesion countries. The only countries to escape this upward trend were Bulgaria, Lithuania, and Romania. Difficulties in collecting social insurance contributions sometimes exacerbate the resulting financial strains. Health care may have become more costly, but the CEECs stand out among transition countries for having maintained or even improved the health status of their population. Life expectancy has improved to between 70.2 and 75.8 years, and so has infant mortality. However, infant mortality remains very high in countries where health spending per capita is lowest (Bulgaria, Latvia, and Romania). The same countries are also exhibiting high death rates from cardiovascular and cerebro-vascular ailments. One important change since the mid-1990s is that the further rise in health costs has generally been associated with higher (formal or informal) out-of-pocket payments on the part of patients, rather than with additional public spending. Indeed, public spending on health has broadly stabilized (relative to GDP) since the mid-1990s. The levels of public spending, however, remain comparatively high in a number of countries (including, for example, the Czech Republic), and the health contribution rates levied to finance them (generally around 13-14 percent of wages) are uniformly high in comparison with those in the EU. The need for efficiency gains remains, therefore, as pressing as ever. Despite sweeping changes to the design of health systems, there is as yet little sign of the expected efficiency gains. In some cases this is because, despite the talk of separating the financing and provision of health services, introducing competition in service provision, and covering the population with mandatory health insurance, there has been considerably less change in the way the sector actually operates. In addition, countries are still struggling to define appropriate payment formula to tie their general practitioners, specialists, and hospitals into an appropriate framework of efficiency-enhancing incentives. In this context, a number of recent developments look promising, including in particular, the effective separation of service providers from ministries of health (beyond a core number of public health and last resort institutions), and their incorporation into various forms of ownership; the spread of mixed provider payment formulas and the first experimentation with diagnosis related groups schemes for the payment of hospitals; the merger of health insurance collection with that of other social security contributions; and the regularization of private payments into a system of well-regulated patient co-payments. An Expenditure Strategy for Growth and Convergence Expenditure policies can make a decisive contribution to the CEECs' objective of growth and convergence with the EU. In keeping with the trends observed within the EU (including within the cohesion group) as well as the OECD at large, a first element of such a strategy would be to curb aggregate expenditure levels, as most countries have indeed suggested in xx their PEPs. A first rationale for such an approach is to leave more resources in the hands of a more efficient private sector and attain a more rapid factor productivity growth. Of equal importance, such a strategy would create the fiscal space needed to reduce the burden of labor taxation, thereby activating resources (that is, labor) that are currently sitting idle. At present, the burden of labor taxation in the CEECs (at about 75 percent of wages) exceeds even the high levels that saddle EU labor (at an average rate of 53 percent) and is even further out of line with the levels observed in the broader OECD. It is difficult to see how the CEECs could live up to the aspirations of the Lisbon Strategy that they have recently espoused, of turning the enlarged EU into "the most competitive economy in the world with social cohesion and full employment of the 70 percent of the adult population" without bringing those tax wedges down. A primary tool for curbing expenditure should be to rein in subsidies and transfers. Currently, this category of expenditure makes up as much as 60 percent of total primary expenditure in the Czech Republic and the Slovak Republic, as compared to 40 percent of primary expenditure in the EU. Even at the other end of the range, Lithuania still devotes relatively more of its expenditure envelope to subsidies and transfers (38 percent) than the average cohesion country (35 percent). Indeed, despite years of progress in tightening budget constraints, considerable scope remains in a number of countries to cut back on enterprise subsidies. This is the case with the Czech Republic and the Slovak Republic where enterprise subsidies account for more than 12 percent of GDP, but even Estonia, Hungary, Latvia, and Romania still spend more than twice as much relative to GDP (about 4 percent of GDP) as the average EU country (1.5 percent of GDP) on subsidizing various categories of enterprises. Part of the answer lies in completing bank restructuring and expediting large-scale enterprise privatization. In some cases (for example, coal mines) the initial restructuring costs may be high. Provided the restructuring strategies are well tailored to the ultimate privatization of the enterprises, such up-front costs are likely to be worth paying, for the sake of the longer-term fiscal benefits that would arise from taking ailing firms off of the public books. A second source of fiscal relief would come from the implementation of the acquis communautaire in sectors such as power, gas, and transport to the extent that it protects government from the ever-present temptation to provide state aid. Similarly, much more could also be done to curtail social transfers. In Bulgaria, the Czech Republic, and Hungary, social transfers exceed the levels observed in the cohesion countries; in Slovenia and Poland, they are even higher (relative to GDP) than among EU members. Some of the CEECs are acting, apparently successfully, to address this problem (witness the 2-3 percentage points of GDP reduction in social transfers in Poland and Hungary between 1995 and 2000). In the absence of systemic action, the upward trend appears inexorable in the Czech Republic (where spending on social transfers rose by more than three percentage points of GDP between 1995 and 2000), and the Slovak Republic might soon follow suit. Therefore, the time has come even for those countries that have avoided any pension system crisis to move ahead with more systemic reform if they wish to secure old-age protection over the long run while improving labor market incentives in the short run. xxi Health care reform is another way of pursuing the latter objective. To make up for persisting inefficiencies in the sector, current health insurance contribution rates are set far higher than those prevailing within the EU. Above and beyond the direct benefits it would bring to the sector, more decisive action on health care reform, by creating room for cutting back the payroll tax, would therefore also contribute indirectly to the overarching growth and convergence the CEECs have set out for themselves. We have highlighted ways in which key expenditure programs could be redirected toward growth objectives. The education sector, for instance, could make a better contribution to CEECS' integration into the global knowledge economy if countries took advantage of the passing of a "baby bust" wave to downsize the staffing and facilities of basic and vocational education and shift the resources toward other educational inputs as well as higher or more generic levels of education. Transport networks can also play a key role in facilitating growth. But their impact will be more powerfully felt if investmient decisions are guided by economic rather than by other considerations. In practice, for many countries this would imply prioritizing maintenance expenditures and phasing in new motorway construction only as the underlying traffic justifies. Similarly, the rates of return on environmental programs can vary within a considerable margin. In some cases, they can actually be quite high. Estimates presented in this report suggest that clean air (and to a lesser extent clean water) measures would yield the highest economic benefits in relation to costs, but they raise questions about the amount of investment in waste management that would be economically viable. These indications are still tentative, and the merit of each individual project will need to be ascertained on a case-by-case basis. But the cause of growth and convergence will be better served if those projects showing the highest return receive first priority. In conclusion, the general thrust of expenditure strategies put forward in the PEPs appears both appropriate and at least theoretically feasible. Factors related to a country's political economy and to the institutional framework of public resource management will undoubtedly play a determining role in framing actual policy choices. xxii 1. EXPENDITURE TRENDS AND CHALLENGES In their Pre-Accession Economic Programs (PEPs) most Central and Eastern European countries (CEECs) are putting forward medium-term fiscal strategies to improve budget balances and cut back taxation levels. The corollary is a need to bring down public expenditure by an average of 2.5 percent of gross domestic product (GDP) over the period 2000-2004. The National Programs for the Adoption of the Acquis, on the other hand, figure that accession may entail additional expenditures averaging 3.5 percent of GDP. Reconciling these two sets of objectives can only be done through deep cuts in regular expenditure programs. In this chapter we will seek to identify the scope for such public expenditure restructuring. We will discuss the fiscal objectives the CEECs have set for themselves in various official documents and will compare those objectives to recent trends in public expenditure aggregates across CEECs. At an aggregate level, it appears at first that there is little room for public expenditure restructuring and rationalization. This observation motivates a further inquiry into the detailed composition of expenditure across the European Union (EU) accession group and in selected comparator countries. The picture that emerges from this closer look is that instead of the initial impression of inertia and uniformity, there is actually considerable variation in spending levels across candidate countries, and that this variation provides prima facie evidence (contrary to the spirit of inevitability that often pervades policy debates on expenditure reform) that different policy options are at least conceivable. FISCAL OBJECTIVES OF THE CEECS Most PEPs plan to reduce overall budget spending in the medium term (except those of the Czech Republic and Romania, which envisage only containing the growth of public expenditure relative to GDP) (table 1.1). The planned public expenditure cuts vary from country to country (ranging from about 1 percent of GDP between 2000 and 2004 in Slovenia to about 4 percent of GDP in Bulgaria), but the objective is the same: to make room for tax reduction while progressing toward the fiscal objectives of the EU's Stability and Growth Pact. PEPs are, however, often less explicit as to how expenditure cuts would be achieved. - 1 - Expenditure Policies Toward EU Accession Table 1.1 Pre-Accession Fiscal Programs of the CEECs, 2000-04 (percent of GDP) Revenue Expenditure Balance 2000 2004 2000 2004 2000 2004 Bulgaria 43.5 38.9 44.5 40.4 -1.0 -1.5 Czech Republic 40.6 41.3 43.8 45.6 -3.2 -4.3 Estonia 38.9 37.7 39.6 38.1 -0.7 -0.4 Hungary 45.9 43.1 48.9 45.1 -3.0 -2.0 Latvia 30.0 27.0 32.7 27.7 -2.7 -0.7 Romania 31.5 34.0 35.5 37.0 -4.0 -3.0 Lithuania 30.2 28.7 33.0 30.1 -2.8 -1.4 Poland 39.6 36.9 42.7 40.5 -3.1 -3.6 Slovak Republic 35.3 29.3 38.7 31.8 -3.4 -2.5 Slovenia 42.8 42.8 44.1 43.2 -1.3 -0.4 Average CEECs 37.8 36.0 40.4 38.0 -2.5 -2.0 EU 47.0 45.8 1.2 Source: CEECs, Pre-Accession Economic Programs, 2001. The CEECs estimate that EU accession will cost them annually on average 3.5 percent of GDP (table 1.2). Bulgaria's estimation is on the higher end, with financing needs for compliance at 7.6 percent of GDP (of which 3.3 percent would have to be financed from the state budget). However, leaving Bulgaria aside, the average "cost of EU accession" still works out at about 2.5 percent of GDP.' Table 1.2 Fiscal Cost of EU Accession and Sources of Financing: National Estimates (annual average, percent of GDP) Local O.W. state Other local EU Other Total financing budget sources financing financing Period Bulgariaa 7.6 3.6 3.3 0.3 2.2 1.9 2000-06 Czech Republic 3.2 2.4 1.1 1.2 0.3 0.5 2001-02 Hungary 2.1 1.2 1.2 0.0 0.5 0.4 2000-02 Latvia 3.3 1.6 1.4 0.2 1.1 0.6 2002-02 Lithuania 1.7 1.0 0.6 0.4 0.4 0.2 2001-09 Poland 1.0 0.4 0.2 0.2 0.6 0.0 2000-03 Romania 3.8 1.2 1.2 0.0 0.9 1.6 2001-04 Slovak Republic 3.6 1.3 1.3 0.0 0.9 1.4 2000-02 Slovenia 2.6 2.2 2.2 0.0 0.4 0.0 1999-03 Average CEECs 3.2 1.6 1.4 0.3 0.8 0.7 a. Excluding private sources. Source: CEECs, National Program for Adopting the Acquis. The 2000 GDP was used to calculate financial needs for EU accession in terms of GDP. I Although EU pre-accession funds are expected to be an important source for EU-related expenditures, 50-70 percent of financing needs would be funded from domestic, mainly public, sources. The CEECs expect that EU pre- structural funds and Phare programs would cover about 20-25 percent of the costs, and 10-20 percent of financing would come from other sources. -2- Expenditure Trends and Challenges Taken at face value, these figures would imply that the CEECs are contemplating cuts in their regular expenditure programs on the order of'5.5 percent of GDP on average, and even twice as much in countries such as Bulgaria and the Slovak Republic. Of course, the figures presented in table 1.2 may well have an upward bias.2 But even assuming that EU accession costs would be only half as high, the implied expenditure adjustment would still be quite sizable (at 4.2 percent of GDP, on average). RECENT TRENDS IN GENERAL GOVERNMENT EXPENDrTURE Against these estimated costs it can still be seen that consolidated general government spending levels in the CEECs have actually not changed much in recent years (table 1.3). Ordy in Estonia, Hungary, and the Slovak Republic did the expenditure adjustment exceed 3 percent of GDP between 1995 and 2000. Plus, those cuts were almost offset by expenditure expansion in other countries, so much so that, considering the CEECs as a whole, public expenditure hardly budged (relative to GDP) between those two years. Table 1.3 Consolidated General Govermnent Expenditure (Excluding Net Lending), 1995-2000 (percent of GDP) 1995 1996 1997 1998 1999 2000 Lithuania 31.6 30.9 31.7 35.7 37.5 32.9 Latvia 38.9 40.2 38.3 42.4 42.7 38.8 Estonia 41.3 40.5 38.0 38.8 41.8 38.1 Czech Republic 43.0 -42.2 41.7 41.0 42.2 44.3 Poland 44.9 44.6 44.7 42.7 43.6 43.0 Romania 34.2 33.5 33.2 34.7 34.9 35.1 Hungary 52.3 48.5 49.0 47.6 47.2 46.2 Bulgaria 44.0 51.3 36.3 36.1 39.1 42.0 Slovenia 43.2 42.4 43.3 43.8 44.3 44.2 Slovak Republic n.a 45.5 43.9 41.9 40.3 43.1 Average CEECs 41.5 42.0 40.0 40.5 41.4 40.8 Average cohesion countrnes 45.0 43.7 42.3 41.5 41.8 40.7 Average EU-15 51.4 51.1 49.4 48.4 47.9 45.8 Average OECD 44.4 43.8 42.5 42.1 41.8 40.6 Average OECD, excl. Continental Europe 36.1 35.5 34.4 34.5 34.2 33.7 Note: The data used in this study come from the IMF's Government Finance Statistics (GFS) database. The GFS data were consolidated following the GFS 86 methodology. Sources: European Commission General Government Data (ESA95); OECD Government Finance Statistics. 2 First, the implhed expenditure choices may not always reflect the available least cost options. Second, the extent to which those "accession-related" expenditures indeed represent additional expenditures or would have taken place irrespective of EU accession is a matter of debate. It is difficult, for instance, to draw the line between investments needed for EU accession, and overall development needs. Finally, we cannot exdude the possibility that some of the underlymg estimates would have been inflated in a bid for larger state budget or pre-accession funds. - 3 - Expenditure Policies Toward EU Accession Furthermore, much of the expenditure containment that has taken place can be traced to a decline in interest payments, owing to a reduction in the overall debt burden. Whether the same relief can be relied upon in the future to bring down aggregate spending is a different matter. Opposing the general downward trend, interest expenditures are already on the rise in Romania, Lithuania, and the Slovak Republic as their govermment debts have increased, and are bound to follow the same path in the Czech Republic (as the government absorbs the costs of bank restructuring) and Poland (as result of the current widening of the fiscal deficit, see table 1.4). Table 1.4 Central Government Debt and Interest Payments (percent of GDP) Central government debt Interest Implicit interest rate 1995 2000 1995 2000 1995 2000 Lithuania 17.8 28.1 0.4 1.8 2.3 6.3 Latvia n.a 13.2 1.2 1.1 n.a 8.5 Estonia 8.6 5.3 0.2 0.4 2.5 7.5 Czech Republic 15.3 17.0 1.2 1.1 7.8 6.4 Poland 54.3 38.9 4.5 2.7 8.3 6.9 Romania n.a 31.6 1.4 4.9 n.a 15.5 Hungary 86.3 56.9 9.3 6.2 10.7 11.0 Bulgaria 120.6 74.9 14.6 4.3 12.1 5.7 Slovenia n.a 20.8 1.2 1.5 n.a 7.2 Slovak Republic 22.8 32.8 n.a 3.2 n.a 9.8 Average CEECs 46.5 32.0 3.8 2.7 7.3 8.5 EU 64.2 3.8 5.9 Sources: World Bank; IMF Government Finance Statistics. Even if they occur, further reductions in the interest bill are unlikely to be large enough on their own to bring about the targeted expenditure retrenchment. An easing in interest rates may help. But the primary deficits that most CEECs are running (with the notable exception of Hungary and Bulgaria) will operate in the opposite direction. Unfortunately, the record of curbing primary expenditure is at best mixed. Once interest payments are netted out, the picture that emerges is one of expenditure inertia. On average, primary expenditure remained steady at about 38 percent of GDP over the period. Estonia did compress spending on goods and services, while Hungary and the Slovak Republic managed to cut back on current transfers. But Bulgaria used the savings it made on interest payments in recent years to lift overall public spending with major increases in allocations for capital budgets and current transfers (table 1.5). -4 - Expenditure Trends and Challenges Table 1.5 Consolidated General Government Primary Expenditure, 1995-2000 (percent of GDP) 1995 1996 1997 1998 1999 2000 Lithuania 31.2 29.9 30.8 34.5 36.0 31.2 Latvia 37.8 38.5 37.3 41.6 42.0 37.7 Estonia 41.1 40.1 37.5 38.3 41.4 37.7 Czech Republic 41.8 41.1 40.5 39.8 41.2 43.2 Poland 40.4 40.9 41.3 39.5 40.5 40.3 Romania 32.8 31.8 29.4 30.0 29.6 30.2 Hungary 43.1 40.2 39.2 39.8 39.8 40.1 Bulgaria 29.4 31.0 27.9 31.6 35.2 37.7 Slovenia 42.0 41.2 42.1 425 42.9 42.7 Slovak Republic n.a 43.4 41.9 39.1 37.1 39.9 Average CEECs 37.7 37.8 36.8 37.7 38.6 38.1 Average cohesion countries 38.0 37.3 36.9 36.8 37.6 36.7 Average EU-15 46.0 45.6 44.4 43.8 43.8 42.0 Average OECD 39.0 38.6 37.8 37.6 37.6 36.6 Average OECD, excl. Continental Europe 31.7 31.3 30.6 30.8 31.0 30.6 Sources: European Comnission General Govermment Data (ESA95). OECD Governmnent Finance Statistics. Before pursuing further discussion, it is useful to remember that overall public spending in the CEECs is around the level in the so-called "cohesion countries" (Greece, Ireland, Portugal, and Spain),3 and is considerably below the average of the EU-15. This would seem to indicate that perhaps CEEC expenditure patterns have already settled around benchmark levels. As is often the case, however, averages conceal wide variations across countries. In 2000, primary expenditure varied from around 43 percent of GDP in the Czech Republic and Slovenia to 40 percent in Hungary and Poland to only slightly more than 30 percent in Lithuania and Romania. Similarly, primary expenditure levels in cohesion countries ranged from 31 percent to 42 percent of GDP in that year. Some of this variation is associated with differences of income. This is in line with the common statistical observation that the demand for government services (or the capacity to fulfill the demand) expands when income rises (figure 1.1). 3 Local government spending mncreased as a percentage of GDP in a weighted average EU accession country, but the direction of intergovermnental relations is not clear for every country. A significant diversity in shares of local government spending in total expenditure (from 36 percent in Poland to 7 percent in the Slovak Republic and 11 percent in Romania) also shows that expenditure mandates among different levels of government in most countries are not yet clearly defined. 5- Expenditure Policies Toward EU Accession Figure 1.1 General Government Expenditure and Income per Capita in the EU and the CEECs, 1999 65- 0 60 *Sweden 0 Denmark 55 - Austria * France w~~~~~~~~~~~~~~ *Germance 50 -kl ~ Iir _ ~~~~~~~~*HungaryGemn = 45 - Greece p% Netheriand i45- aa Latvia * Slovenb IU * !Dceh Rep.* 40 - _ pain *United K6gdom Bulgaria 4+omania Slovak Rep C0 Lithuania teland 35- , , , , a, 0 5000 10000 15000 20000 25000 30000 GNI per capita at PPP Source: World Bank. As Figure 1.1 shows, however, there is considerable variation around that trend line. Countries with similar levels of income per capita sometimes have very different levels of government expenditure relative to their GDP: the Czech Republic (42 percent of GDP) and Hungary (47 percent of GDP), for instance, or Italy (49 percent of GDP) and the United Kingdom (40 percent of GDP). A statistical observation is not the same thing as a policy rule. When designing their fiscal strategies, the CEECs may have noted that the normal relationship between income and government size is actually inverted in the case of the cohesion countries and that the country in which government is the smallest-Ireland-is also the one in which income per capita has risen most rapidly since the mid-1980s. Indeed, the general trend of government expenditure across the Organization for Economic Cooperation and Development (OECD) has been downward (figure 1.2). The trend has been most pronounced within the EU (where general government outlays shrank by 5 percentage points of GDP between 1995 and 2000). The relative decline was perhaps less outside of continental Europe, but the starting point was also considerably lower, bringing into even sharper contrast the relative performance of the CEECs. By 2000 general government outlays in the CEECs exceeded the levels observed among non-continental European members of the OECD by a full six percentage points of GDP. -6- Expenditure Trends and Challenges Figure 1.2 Consolidated General Government Outlays in the CEECs and the OECD, 1995-2000 (percent of GDP) 50.0 45.0 = = = = = = 40.0- 35.0- 30.0 - 1995 1996 1997 1998 1999 2000 CEEC EU - -OECD - OECD non-continental Europe Note: OECD non-continental Europe comprises of Australia, Canada, Iceland, Ireland, Japan, Korea, New Zealand, United Kingdom and United States. Source: OECD. DIFFERENCES IN EXPENDrrURE COMPOSITION An even greater variation can be observed when government expenditure is broken down into economic and functional categories (tables 1.6 and 1.7), confirming the view that far from being preordained, expenditure levels and the mix of expenditures are subject to a fair degree of policy discretion. For any particular item of the Government Finance Statistics (GFS) classification, the levels of expenditures (relative to GDP)4 vary by at least a factor of two.5 Current expenditures, for instance, differ by as much as 12 percent of GDP between the top and the bottom ends of the range (table 1.6). The case of defense expenditure is particularly telling (table 1.7). All concerned countries are either preparing for integration into NATO and/or have recently joined it. Still, despite this similarity of circumstances, defense spending varied in the range of 0.8 percent of GDP in Latvia and 3 percent of GDP in Bulgaria in 2000 (this compares with an EU average of 2 percent of GDP). 4 Average over the period 1995-2000. 5 Except for expenditures on general public services and on public order, which vary only by a factor of 1.5 and 1.8, respectively. -7 - Table 1.6 Consolidated General Government Expenditure (Excluding Net Lending) by Economic Type, 1995-2000 (percent of GDP) Other Subsidies Expenditure Wages purchases of and other Total Current on goods and goods and current Capital Capital expenditure expenditure and services salaries services Interest transfers Expenditure Investment transfer Lithuania 33.4 29.8 17.7 7.8 10.0 1.1 10.9 3.6 3.0 0.6 Latvia 40.2 37.0 16.8 7.7 9.0 1.1 19.1 3.3 1.6 1.7 Estonia 39.8 35.5 20.1 5.4 10.8 0.4 14.9 4.3 4.1 0.2 Czech Republic 42.4 36.5 19.3 7.7 11.6 1.1 16.1 5.9 3.9 2.0 Poland 43.9 40.7 15.8 7.4 8.4 3.4 21.5 3.2 2.8 0.4 Romania 34.3 30.1 11.8 5.5 6.3 3.4 14.7 4.1 2.5 1.6 Hungary 48.5 42.6 15.1 7.6 7.5 8.1 19.4 5.9 4.1 1.8 Bulgaria 41.5 38.1 15.3 4.9 7.5 9.3 13.5 3.3 3.0 0.3 Slovenia 43.5 39.0 17.9 9.4 8.5 1.3 19.8 4.5 2.7 1.8 Slovak Republic 42.9 37.3 16.6 9.6 7.0 2.6 18.0 5.6 4.0 2.2 Average CEECs 41.0 36.7 16.6 7.3 8.7 3.2 16.8 4.4 3.2 1.4 Average Cohesion countries 42.5 37.3 16.7 11.3 5.4 5.3 13.7 5.2 3.4 1.8 Average EU-15 49.0 45.0 20.3 10.7 9.6 4.7 18.3 4.0 2.3 1.7 Average OECD 42.3 40.0 18.5 10.8 7.8 4.7 16.8 3.2 3.2 0.0 Av. OECD, excl. Cont Europe 34.7 32.6 16.9 9.6 7.3 3.8 11.6 3.4 3.2 0.2 Sources: IMF Government Finance Statistics; European Commission General Govermnent Data (ESA95). OECD Government Finance Statistics. Table 1.7 General Govermnent Expenditure (Excluding Net Lending) by Function, 1995-2000 (percent of GDP) Social Housing Miscellaneous General Public security and Recreation (including public order and and community and Economic statistical Total services Defense safety Education Health welfare amenities cultural services Interest adjustment) Lithuania 33.4 1.6 0.8 2.3 5.7 4.3 10.5 1.0 1.0 4.3 1.1 0.8 Latvia 40.2 3.3 0.8 2.3 6.5 3.9 14.8 1.9 1.3 4.5 1.1 -0.2 Estonia 39.8 3.1 1.4 2.7 6.9 5.7 11.1 1.5 2.2 4.7 0.4 -0.1 Czech Republic 42.4 2.9 1.8 2.2 4.6 6.6 13.3 3.4 1.0 6.1 1.1 -0.8 Poland 43.9 2.4 1.5 1.8 5.6 4.5 19.9 3.5 0.8 2.9 3.4 -2.6 Romania 34.3 1.3 2.0 1.5 3.3 3.1 10.1 1.8 0.5 6.1 3.4 1.2 Hungary 48.5 3.4 1.1 1.6 4.8 4.4 15.4 1.5 1.5 5.5 8.1 2.4 Bulgaria 41.5 2.7 2.7 2.0 4.0 3.6 11.6 1.5 0.7 5.2 9.3 -1.8 Slovenia 43.5 3.9 1.2 1.6 5.6 5.7 17.8 1.6 1.2 5.0 1.3 -1.3 Slovak Republic 42.9 3.4 2.0 2.0 4.1 5.8 13.7 2.0 1.0 7.2 2.6 -0.7 Average CEECs 41.0 2.8 1.5 2.0 5.1 4.8 13.8 2.0 1.1 5.1 3.2 -0.3 Average Cohesion Countries 42.5 2.0 2.4 0.8 5.0 5.1 10.1 1.3 0.6 3.8 5.3 Average EUa 49.0 2.8 2.0 1.0 5.3 6.6b 18.3 1.8 0.8 3.8 4.7 Average OECD 42.5 2.7 2.1 1.3 5.5 5.7 16.1 1.7 0.9 3.5 4.0 Av. OECD, excl. Cont. Europe 34.7 2.4 2.1 1.4 5.5 5.4 11.3 1.6 0.9 2.8 4.5 a/ Excluding Luxembourg. b/ As a percentage of GNI. Sources: European Commission General Government Data (ESA95), OECD Goverrument Finance Statistics, World Development Indicators 2001. Expenditure Policies towards EU Accession A closer inspection of these expenditure categories reveals a number of interesting, and sometimes counterintuitive observations, including the following: * The CEECs spend less than member countries on wages and salaries; their social sectors, however, are often overstaffed. * The CEECs are already catching up with (and sometimes exceeding) the high levels of public investment observed among cohesion countries, raising concerns about the quality of the project pipeline and the possible displacement of operations and maintenance expenditure. * The CEECs are devoting the larger share of their public resources to funding transfers and subsidies. Early reformers are beginning to see the benefits of their efforts to reduce this kind of spending, but pressures are mounting for the later reformers. * The CEECs have, in some cases, not yet managed to rein in state aid (broadly speaking), in part because of the ongoing cost of bank restructuring. Wages and Salaries The CEEC governments do not appear to spend excessively on wages and salaries. The figures vary considerably from country to country, but CEECs typically spend less on wages and salaries than cohesion countries (or the EU in general), though a number of the candidate countries have a larger number of government employees than the cohesion countries; none of the candidates matches the EU average in terms of government employment (table 1.8). The social sectors, however, are generally overstaffed. Candidate countries typically maintain smaller central and local bureaucracies than member countries but employ larger numbers of education and health personnel than countries in the cohesion group (and the EU in general). Education systems in most CEECs have failed to adjust to the secular decline in their school age population (table 1.9) and still have a long way to go to restructure their health systems.6 Thus, in 2000 education and health expenditures were relatively higher in a number of CEECs (the Baltics, Poland, and Slovenia; the Czech Republic, Estonia, the Slovak Republic, and Slovenia, respectively) than among cohesion countries. Other CEECs have merely managed to compress wage levels, sometimes to such an extent as to trigger the spread of "informal payments" for social services. 6 In a country such as the Slovak Republic the number of teachers expanded through the 1990s even as the school age population contracted, and the numbers of doctors and nurses, though already comparatively higher than in the OECD, continued to rise. - 10- Expenditure Trends and Challenges Table 1.8 Government Employment and Expenditure on Wages and Salaries, 1996-2000 (percent of population and percent of GDP, respectively) Government wage bill Central Local (percent of Total government governments Education Health GDP) (percent of population) Lithuania 8.2 1.3 0.5 4.0 2.5 7.9 Latvia 8.4 1.9 4.2 1.5 0.9 7.6 Estonia 6.4 1.3 0.2 2.9 2.0 6.5 Czech Republic 8.1 4.8 2.3 0.7 0.3 7.7 Poland 5.2 0.3 0.4 2.1 2.4 7.3 Romanua 4.2 0.5 0.5 1.9 1.3 5.3 Hungary 7.6 1.5 1.6 2.3 2.2 7.4 Bulgaria 6.4 1.2 0.6 2.8 1.8 4.9 Sloverua 3.8 1.4 0.2 1.3 0.9 9.5 Slovak Republic 6.4 0.5 0.3 3.2 2.4 9.6 Average CEECs 6.5 1.5 1.1 2.3 1.7 7.4 Average Cohesion Countries 6.9 3.2 1.4 1.4 0.9 11.3 Average EU 8.9 2.6 3.2 1.8 1.4 10.7 United States 10.4 1.0 5.6 3.3 0.5 9.2 Sources: World Bank Administrative and Civil Service Reform Thematic Group; OECD public sector employment data; CEEC governments. Table 1.9 Growth of the Age 0-14 Population (annual percent change) 1980-1998 1998-2015 Lithuania -0.4 -1.8 Latvia -0.7 -3.0 Estonua -1.0 -2.3 Czech Republic -1.6 -2.0 Poland -0.4 -1.6 Romania -1.8 -1.8 Hungary -1.6 -1.6 Bulgaria -1.9 -2.4 Slovenia -1.6 -1.7 Slovak Republic -0.8 -1.4 Source: World Bank, World Development Indicators 2000. Indeed, a lower wage bill is not necessarily an objective per se. As Figure 1.3 indicates, it may also be associated with a greater perception of corruption. But reducing excess staffing may create room for improving both pay conditions and the quality of services. - 11 - Expenditure Policies Toward EU Accession Figure 1.3 Govermnent Wage Bill and Perception of Corruption 2.0- 3.0- o 4.0- 0. 0 6.70- 07.0 0 LI 8.0 9.0 10.0 2 2.5 3 3.5 4 4.5 5 5.5 6 Wage Bill Note: Corruption Index indicates perception of corruption where I is greatest perception of corruption and 1o is least perception of corruption. Consolidated General Government Wage Bill as percentage of GDP. Sources: IMF Government Finance Statistics; Transparency International. Public Investnent Taken as a group, the CEECs do not appear to be affected by any underinvestment in public infrastructure. At over 4 percent of GDP on average, the level of capital expenditure in the CEEC is close to that in the EU as well as to the levels recorded in the cohesion countries in the early 1990s. Poland, Lithuania, and Romania might be on the low side (at 2.5-3 percent of GDP). But by 2000, countries such as the Czech and Slovak Republics (about 6 percent of GDP) and Hungary (about 7 percent of GDP) were spending relatively more on capital expenditure than the levels of any of the cohesion countries, even though the cohesion countries had the support of the EU structural funds, and Bulgaria and Slovenia had caught up with the lower end of the cohesion country range (Spain's 4.6 percent of GDP). Clearly, for these countries the key question was becoming not how much but how good that capital expenditure was. Operations and Maintenance A related concern is that the same countries that are spending more on capital expenditure are also devoting comparatively less to operations and maintenance (witness, for instance, the 5 percent of GDP spent on operations and management (O&M) in the Czech and Slovak Republics), which raises questions of whether their overall expenditure mix is appropriate. Nevertheless, it remains the case that the average spending on O&M in the EU accession group is high when compared with the levels observed within the cohesion group. Unfortunately, this should not necessarily be interpreted as a sign that public services are better run. Higher spending may also reflect an inefficient consumption of inputs (for example, energy consumption in low energy-efficiency facilities) or higher unit prices resulting from - 12 - Expenditure Trends and Challenges insufficiently competitive public procurement. These different factors unfortunately cannot be separated on the basis of aggregate data. Subsidies and Transfers Overall subsidies and transfers continue to account for the lion's share of government expenditure in the CEECs. The peaks are reached in the Czech and Slovak Republics where they absorbed 60 and 66 percent of primary expenditure in 2000, respectively. Only in Lithuania was that ratio less than 40 percent. Taken as a whole, candidates spend about the same percentage of GDP on subsidies transfers as the member countries, although they spend substantially more (an average of nearly five percent of GDP) than the cohesion countries. Averages are, however, somewhat misleading. Take, for instance, welfare payments. In 1995 the CEECs were essentially polarized in two groups, with a group of high spenders (Hungary, Poland, and Slovenia) on the one hand devoting 18-20 percent to social welfare, and the rest spending around 10 percent of GDP. By 2000, the picture had changed with two groups of countries converging toward the middle of the range as high spenders (Hungary, Poland) were beginning to experience the benefits of pension reform while spending was drifting upward among some of the former low spenders (the Czech Republic, Bulgaria, and, to a lesser extent, Estonia and Lithuania). Only the Slovak Republic and Slovenia appeared to succeed in keeping a lid over their social transfers. The differences in social transfers, however, do not necessarily reflect the countries' living standards. Poland is saddled with a similarly high poverty rate despite spending almost twice as much as Estonia on social security and welfare programs as a percent of GDP. Conversely, the Slovak Republic has two-and-a-half times the unemployment rate of Hungary, and spends two-thirds as much on social security and welfare, but has only half the poverty rate of Hungary. In contrast to what was observed regarding welfare payments, governments appear to have been unable to affect the levels of state aid.7 By 2000 current and capital subsidies to enterprises still exceeded the 1995 levels across the region, indicating the continued struggle to break with the past. This is not for the lack of available options: These levels ranged all the way from 0.6 percent of GDP in Lithuania to 7.3 percent of GDP in the Slovak Republic, the widest cross- country variation in any expenditure item. 7 Estimated as the difference between "Current Subsidies and Transfers" plus "Capital Transfers" (table 1.6) and "Social Security and Welfare" (table 1.7). - 13 - Expenditure Policies Toward EU Accession Table 1.10 Poverty Incidence and Welfare Payments, 1995-2000 Poverty incidence (percent of Expenditure on social security and population at US$4.30/day) welfare (percent of GDP) Lithuania 22.5 10.5 Latvia 34.8 14.8 Estonia 19.3 10.9 Czech Republic 0.8 13.2 Poland 18.4 19.8 Romania 44.5 10.1 Hungary 15.4 15.4 Bulgaria 18.2 11.6 Slovenia 0.7 17.8 Slovak Republic 8.6 10.7 Average CEECs 18.3 13.5 Sources: Poverty incidence (latest available date) from Revenga, A, et al. 2000. Making Transition Workfor Everyone. Washington, D.C.: World Bank; fiscal data: IMF Government Finance Statistics, except for Hungary and Romania, national data. CONCLUSIONS There is evidence that the scope exists to carry out the type of expenditure reform implicit in the CEECs' PEPs. How to bring it about is a different matter. We now turn to the practical experience of dealing with the three key expenditure challenges confronting the candidates: (a) liquidating the legacy of transition, (b) facilitating the countries' successful integration into the EU, and (c) finding new ways to ensure social protection. - 14- 2. LIQUIDATING THE LEGACY OF THE PAST The incapacity of most CEECs to enforce financial discipline on enterprises, at the beginning of the transition, has left a painful and a burdensome legacy to this day. Banking problems were only one manifestation of the underlying enterprise problems. In addition to receiving "soft" loans from the banking sector, unviable enterprises were staying afloat through arrears to the state (most prominently arrears in taxes and in social security contributions) as well as arrears to other enterprises (in particular, to public utilities). For many CEECs, liquidating this legacy continues to top the list of fiscal priorities. A key issue facing the CEEC governments is how to minimize fiscal costs while liquidating the legacy. This will require a threefold strategy aimed at stemming financial losses and cleaning up balance sheets while minimizing the risks that similar problems will recur in the future. The prospect of EU accession has proved a helpful catalyst for action. Not only does it provide a welcome degree of urgency to the task of addressing outstanding problems (as is necessary if CEECs want to withstand competition in single integrated markets and cope with the additional risks brought about by capital account liberalization), but also the adoption of the acquis communautaire is providing the necessary framework (for example, in the fields of state aid, energy, transport, and financial sector regulation). RESTRUCTURING BANKS All transition countries-and the CEECs are no exception-have faced serious banking problems at some point over the last ten years or so of transition (table 2.1). When these problems did not lead to full-blown bank runs (as in Bulgaria in 1996-97) they almost always caused severe banking distress. For example, in Romania the resulting nonperforming loans exceeded half of bank portfolios (table 2.1). Table 2.1 Shares of Nonperforniing Loans (percent of total loans, end of period) 1994 1995 1996 1997 1998 1999 2000 2001 Bulgaria 7 13 15 13 12 18 11 4 Czech Republic 37 27 22 20 20 22 19 14 Estonia 4 2 2 2 4 3 2 2 Hungary 20 12 9 5 7 4 3 4 Latvia 11 19 20 10 7 7 5 3 Lithuania 27 17 32 28 13 12 11 8 Poland 34 24 15 12 12 15 16 - Romania 19 38 48 57 59 35 4 3 Slovak Republic 30 41 32 33 44 33 26 16 Slovenia 14 9 10 10 10 9 9 5 Note: Figures for 2001 are projected. They include loans classified as substandard, doubtful, and loss. Sources: European Bank for Reconstruction and Development, 2001. Transition Report, 2001; International Monetary Fund; Central banks. - 15 - Expenditure Policies Toward EU Accession The CEECs have generally made progress in cleaning up bank balance sheets, as can be seen from the generally declining trends of nonperforming loans (table 2.1). While bad loan portfolios still appear large in the Czech Republic, the problem is less severe than it seems because most of the related loans are either "ring-fenced" or fully provisioned. Despite this, the problem has not disappeared. In many countries the numbers only appear smaller because nonperforming portfolios have been carved out and transferred to a specialized debt recovery agency. For instance, bad loans in Czech Republic with a face value in excess of 20 percent of GDP are now within the purview of Konsolidacni Agentura (including loans ring-fenced in the balance sheet of commercial banks).8 How to dispose of them remains a key issue. Before turning to this issue, it is necessary to take stock of how to bring about bank restructuring. What follows are the broad lessons learned: * The earlier the portfolio problems are openly recognized and dealt with, the better the outcome will be. * Successful bank restructuring ultimately requires bank privatization, preferably to strategic investors. * Banks need to be fully recapitalized first9 for bank privatization to be successful. Portfolio Restructuring The experience of the CEECs shows that adequate recapitalization is important not only for successful privatization but for successful bank restructuring in general. Otherwise, there is a high risk that the banking regulator will show excessive regulatory forbearance, or that the new private owners will see no better option than to engage in the same type of reckless behavior as their predecessors. Hungary provides an example in this regard. The first bank restructuring program in 1992-93 did not adequately recapitalize the banks, and a second round of recapitalization was necessary in 1993-94. Another example is Bulgaria, where the 1991-94 operation to clean bank portfolios from bad credits did not provide adequate capitalization so that banking problems worsened, culminating in a crisis in 1996-97. Similarly, in the Czech Republic the initial privatization of an unrestructured bank led to so severe a failure that the bank had to be taken over and re-privatized again (restructured, in this case). Quite apart from the economic costs (in extreme cases financial intermediation can be disrupted and economic growth undermined) banking problems are fiscally costly. Recent estimates indicate, for example, that the restructuring of Czech banks may cost as much as 10 percent of GDP for the period 2001-04. One study that looked at banking crises in the transition economies found that for the period 1991-98 the total fiscal costs of such crises for some CEECs 8 World Bank, 2001, Czech Republic: Enhancing the Prospectsfor Growth with Fiscal Stability. Washington, D.C. 9 World Bank, 2001, World Development Report: Building Institutionsfor Markets. New York: Oxford University Press. - 16- Liquidating the Legacy of the Past ranged from 2 percent of GDP in the case of Estonia to 42 percent of GDP in the case of Bulgaria (table 2.2).10 The bulk of the fiscal costs are direct costs incurred by the government in restructuring banks, the largest share of which is made up of the costs of recapitalizing banks. Recapitalization entails cleaning the balance sheets of banks, with bad loans either being exchanged for government (or government guaranteed) bonds or marked down to market (see the section on bad debt recovery for a discussion of different ways of treating bad loans), and injecting new capital. In addition, in some CEECs, government compensation to depositors in the event of bank liquidations also added to the fiscal costs. Finally, central banks in some CEECs also bore the costs of bank restructuring, most of which consisted of provisions and losses on credit extended to distressed banks. Such costs also eventually find their way into government budgets. Table 2.2 shows the breakdown of total fiscal costs by these three categories for selected CEECs. Table 2.2 Total Fiscal Costs of Bank Restructuring in Selected CEECs, 1991-98 (percent of GDP) Cost to the Government For bank For deposit Cost to the central restructuring compensation bank Total Bulgaria 26.5 3.3 11.8 41.6 Czech Republic 20.6 none 4.8 25.4 Hungary 12.9 none 0.0 12.9 Poland 7.4 none none 7.4 Estonia 1.4 none 0.5 1.9 Latvia 2.5 0.0 0.1 2.7 Lithuania 1.7 1.3 0.2 3.1 Source: Tang, Helena, Edda Zoli, and Irina Klytchnikova. 2000. "Banking Crisis in Transition Economies: Fiscal Costs and Related Issues." Policy Research Working Paper 2484. World Bank, Washington, D.C. Fortunately, these costs do not have to be met at once but can be amortized over time. This is the role of the various forms of bank restructuring bonds (or guarantees) that governments have used in the process. Even so, the orders of magnitudes involved can be sizable. The Slovak Republic expects to pay the equivalent of 0.8 percent of GDP in interest on government bonds issued for bank restructuring. This is in line with the charges incurred earlier by some other CEECs as compared with 1.6 and 1.0 percent of GDP in initial interest payments incurred by Hungary and Slovenia to restructure their banking systems, respectively. Bank Privatization Successful bank restructuring also requires a clean break between the government and the restructured bank to remove the temptation of further political meddling with banking 10 Tang, Helena, Edda Zoli, and Irina Klytchnikova. 2000. "Banking Crisis in Transition Economies: Fiscal Costs and Related Issues." Policy Research Working Paper 2484. World Bank, Washington, D.C. - 17- Expenditure Policies Toward EU Accession decisions. This entails passing majority control to new, preferably strategic, private owners." The CEEC countries that have been forerunners in privatizing banks, in particular to foreign strategic investors (Hungary and Estonia), also have some of the stronger banking systems in the group. The rest of the region is now rapidly emulating them (table 2.3). Foreign ownership can be more effective in insulating banks from pressures, from the enterprise sector or from the government, to extend credit to failed enterprises. Foreign banks can also exert competitive pressures on the domestic banking system12 that can stimulate improvements in efficiency and innovation and can strengthen supervision and regulation. Through these channels foreign banks can help reduce the potential of future banking problems and the associated fiscal costs. A final lesson is that when portfolio problems arise it is better to deal with them early and openly. One reason for the very large size of the Slovak and Czech bailout operations was that the underlying problems with bad loans were allowed to fester for a long period before the governments would deal with them frontally. In the case of the Czech Republic, the problem was compounded by the opacity of the various mechanisms that were initially used.'3 That is, these mechanisms did more to obfuscate the issue (and to conceal its fiscal implications) than to resolve it. With the lessons learned from that experience, the more recent Slovak bailout chose the more transparent procedures adopted by countries such as Hungary and Poland (involving in particular the issuance of recapitalization bonds directly by the government). Table 2.3 Assets of Majority Foreign-Owned Banks (percent of total bank assets) Bulgaria (2000) 74.4 Czech Republic (September 2001) 70.4 Estonia (September 2001) 97.5 Hungary (une 2001) 62.6 Latvia (2001) 62.6 Lithuania (2001) 83.9 Poland (2000) 69.6 Romanma (September 2001) 55.0 Slovak Republic (une 2001) 65.3 Slovenia (March 2001) 5.7 Sources: National banks of Bulgaria, CzechlRepublic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic, and Slovenia. Bad Debt Workout Although banking problems may be subsiding, they have nonetheless left behind a stock of bad debts that in many countries still need to be resolved. Resolving bad debts and recovering assets from enterprises, if done properly, can help alleviate the fiscal costs of 11 World Bank. 2001. World Development Report: Building Institutionsfor Markets. New York: Oxford University Press. 12 Claessens, S., Demirguc-Kunt and H. Huizinga, 1998, ("How does foreign entry affect the domestic banking market." Poltcy Research Working Paper 1918. World Bank, Washington, D.C.) found that foreign entry increased competition in the banking sector in 80 countries for the period 1988-95. 13 Instead of the government assuming responsibility openly, bad portfolios were initially lodged with a variety of special vehicles operating outside of the framework of the state budget with implicit or explicit government guarantees. The situation has been largely regularized over the last few years with the consolidation of bank restructuring programs around the Konsolidacni Agentura. - 18- Liquidating the Legacy of the Past banking problems and can also spur enterprise restructuring and reduce moral hazard. Therefore, the likelihood of future problems in these sectors, problems that could lead to greater fiscal costs down the road, can be reduced. Two main approaches have been tried in the region to collect bad debt. The first approach, the so-called "centralized" approach, involves removing nonperforming loans from banks and transferring them to an asset management company (AMC) or a bank liquidation agency, which then pursues the recovery of the bad loans. In most countries AMCs are state owned for two main reasons. First, it can be difficult to find private investors interested in assuming large amounts of bad assets without government guarantees. Second, from the government's perspective it can be preferable to own such assets rather than to provide guarantees, since the government can benefit from any upward price movements of such assets.'4 The second approach, the so-called "decentralized" approach, entails leaving nonperforming loans on the banks' books and creating workout units within the banks to pursue the recovery of these loans. Both approaches have their advantages and disadvantages.'5 A study covering the CEECs found no clear superiority of one approach over the other in terms of recovery of bad debt.16 Countries that adopted the centralized approach and created AMCs had recovery rates (loans recovered/total face value of bad loans) of 3-5 percent in the Czech Republic, 5 percent in Lithuania, and 16 percent in Hungary. Poland was the only country among those studied that had some success with the decentralized approach, with a bad loan recovery rate of 17 percent. There is as yet no cross-country evidence on the efficacy of the centralized versus the decentralized approach with respect to enterprise restructuring. A recent study that studies seven AMCs around the world found that AMCs have not been effective in spurring enterprise restructuring but have been more effective in resolving insolvent and unviable financial institutions and selling their assets. However, even achieving the latter objective requires certain conditions such as "an easily liquefiable asset (like real estate), professional management, political independence, a skilled resource base, appropriate funding, adequate bankruptcy and foreclosure laws, good information and management systems, and transparency in operations and processes."'7 In these circumstances, an eclectic, multi-track approach may actually have the best results. In the CEECs, Hungary has successfully addressed its workout problems with a mnixed private sector/state strategy involving both centralized and decentralized approaches.'8 This strategy calls for the use of a centralized state workout institution to manage some large and 14 Klingebiel, Daniela. 2000. "The Use of Asset Management Companies in the Resolution of Banking Crisis: Cross Country Experiences." Policy Research Working Paper 2284. World Bank, Washington, D.C.. 15 Klingebiel, Daniela. 2000. "The Use of Asset Management Companies in the Resolution of Banking Crisis: Cross Country Expenences." Policy Research Working Paper 2284. World Bank, Washington, D.C.. 16 Tang, Helena, Edda Zolh, and Irma Klytchnikova. 2000. "Banking Crisis in Transition Economies: Fiscal Costs and Related Issues." Policy Research Workhing Paper 2484. World Bank, Washington, D.C. 17 Klingebiel, Daniela. 2000. "The Use of Asset Management Companies in the Resolution of Banking Crisis: Cross Country Experiences." Policy Research Working Paper 2284. World Bank, Washington, D.C. 18 World Bank. 2000. Czech Republic: Completing the Transformation of Banks and Enterprises. New York: Oxford University Press. - 19 - Expenditure Policies Toward EU Accession complex cases (generally heavy industrial enterprises), and for the same institution to manage additional assets through contracts with private sector asset managers. In addition, the strategy entails having banks work out a certain proportion of bad assets, and having additional assets sold to the private sector on an individual asset or pool basis. The Slovak Republic has recently pursued a similar multi-track approach involving the sale of pools of loans to private investors, the auctioning of individual loans to smnaller investors, the outsourcing of collection through asset/legal management contracts, and, where appropriate after suitable due diligence, the writing off of noncollectible claims. Certain key elements are necessary for successful workouts if an AMC is used. First, for a state-owned AMC it is important that there is a predetermined time limit for its operations. Otherwise, there is a disincentive for an AMC to be successful in collecting bad debt, because that would bring the AMC closer to its date of dissolution. For this reason, a limited life span was introduced for the state-owned bad debt recovery agency in the Czech Republic, the former Konsolidacni Banka (KoB) (recently renamed Konsolidacni Agentura), during its recent restructuring. A self-perpetuating AMC also creates a moral hazard for continuing banking problems, because it provides banks with a dumping ground for future bad loans. Second, the AMC should be given a clear and limited mandate. The Czech case again provides a good example. In addition to being a debt recovery agency, KoB was also initially supposed to serve as a development bank, and a vehicle through which the government could pursue industrial ventures. KoB was also allowed commercial banking and deposit-taking functions, but its high share of bad assets created a tremendous moral hazard for the undertaking of risky activities. For these reasons the recent restructuring of KoB entailed the spinning off of its development banking activities and associated staff, as well as the removal of its banking license. Enhancing the Institutional Framework of Banking Operations Finally, no matter which approach is adopted, strong creditor rights induding effective insolvency (that is, bankruptcy) and collateral systems are important for successful work outs, in terms of recoveries under bankruptcy and also in terms of enhancing the incentives for debtors to negotiate out of court to avoid bankruptcy. However, building an effective bankruptcy system is taking time in many of the CEECs. Thus, the Slovak Republic is focusing first on improving the liquidation mechanisms under the bankruptcy system to speed up the resolution of unviable enterprises and to promote out-of-court restructurings (the latter because of a more credible bankruptcy threat). The Slovak Republic is also working on improving the efficiency of court-led restructuring mechanisms, a more long-term endeavor.19 For similar reasons the Czedc Republic is focusing its attention on improving liquidation procedures and introducing a fast restructuring or reorganization track.20 19 Drawn from World Bank 2001. "Report and Recommendation of the President on an Enterprise and Financial Sector Adjustment Loan to the Slovak Republic." Report Washington, D.C.. 20 For details, see World Bank. 2000. Czech Republic: Completng the Transformation of Banks and Enterprises. New York: Oxford Umversity Press.. A key feature of the fast track procedure is that the opinion of the majority of creditors can be implemented over the objections of the minority class of creditors. This fast track is still subject to approval by the court, and in lhght of the low average development of the judiciary, will initially be restrictive (access to the track will - 20 - Liquidating the Legacy of the Past Another important reason to strengthen the institutional framework for bank operations (that is, the legal framework, prudential regulations, accounting standards, and banking supervision) is to minimize the risk of a recurrence of financial distress. Indeed, those countries among the CEECs that have the lowest shares of nonperforming loans, Estonia and Hungary, are also the countries that have the best regulatory frameworks (according to the European Bank for Reconstruction and Development (EBRD) overall index of financial regulations, table 2.4). If another reason were needed to enhance prudential regulations, then the prospect of completing capital account liberalization by accession (a process already well under way in many CEECs) would provide it. Effective banking regulations -and the capacity to implement them-are indeed essential if countries are to guard against the risks associated with cross- border capital flows.2' Table 2.4 Banking Sector Reform and Financial Regulations Banking sector reform (2000) Financial regulations (2001) Extensiveness Effectiveness Overall Index Index Index Index Bulgaria 3.0 3 3 3 Czech Republic 3.3 3+ 3 3 Estonia 3.7 4 3+ 4- Hungary 4.0 4- 4- 4- Latvia 3.0 3 3 3 Lithuania 3.0 3+ 4- 3+ Poland 3.3 4 3 3+ Romania 2.7 4 3 3+ Slovak Republic 3.0 3 3 3 Slovenia 3.3 4 4- 4. Note: The EBRD indices range from 1 to 4+, with 4+ being the highest. Source: European Bank for Reconstruction and Development. 2001. Transttion Report, 2001. CEECs have generally made great strides toward establishing the necessary formal frameworks. The EU accession process has been beneficial in accelerating the needed legal reforms in the financial sector and providing guidance on the direction of such reforms. Table 2.5 shows that with respect to the capital adequacy ratio all the CEECs have introduced at least the Basle minimum of 8 percent, with half of them requiring even higher capital adequacy ratios. While they are still smarting from the fairly recent financial distress, commercial banks in the CEECs are generally maintaining higher CARs than the Basle rules provide for. Similarly, the CEECs all have in place a minimum capital requirement of at least ECU 5 million (the EU requirement), with some actually exceeding this level. Finally, 8 of the 10 CEECs have introduced International Accounting Standards. depend on assets and liabilities in excess of stated amounts, as well as a certain number of employees). One of the principal objectives of this fast track is to encourage out-of-court negotiations (backed by a more credible bankruptcy threat posed by the improved liquidation procedures). In addition, a fast, bankruptcy-oriented solution for corporate rescues could promote institutional capacity and long-term development within the court system for handling corporate rescues. 21 Wagner, Nancy, and Dora lakova. 2001. "Financial Sector Evolution in the Central European Economies: Challenges in Supporttng Macroeconomic Stability and Sustainable Growth." IMF Working Paper. International Monetary Pund, Washington, D.C. - 21 - Expenditure Policies Toward EU Accession In several CEECs however, further improvements need to be made on the implementation of financial regulations. As can be seen in table 2.4, the effectiveness of financial regulations (which measures the extent to which financial legal rules are clear, accessible, and adequately implemented, both judicially and administratively) typically lags behind their extensiveness (which measures whether banking and capital market legal rules approach minimum international standards).22 Improving the implementation of financial regulations-which implies improving the quality of banking supervision-will require the upgrading of technical skills and the independence of the supervisory body. Table 2.5 Key Elements of the Banking Institutional Framework Minimum Capital Date capital adequacy adequacy International ratio (percent) ratio (percent) Accounting Minimum capital (year introduced) (2001, latest) Standard in force requirementa Bulgaria 12(1999) 32.0 1998 ECU 5.1 bil.; 10 bil. BGL Czech Republic 8 (1996) 15.2 2000 ECU 14.9 mil.; 500 mil. CZK Estonia 10 (1997) 13.6 1994 ECU 5 mil. Hungary 8 (1993) 15.1 1996 ECU 11.7-0.078 mil.; 3 bil. -20 mil. HUF Latvia 10 (1999) 15.0 1992 ECU 5 mil. Lithuania 10 (1997) 15.5 1996 ECU 5 mil. Poland 8 (1993) 1994 ECU 5 mil. Romania 12 (1999) 26.9 NA ECU 5.3 mil.; ROL 150 bil. Slovak Republic 8 (1996) 19.9 NA ECU 11.4 mil.; 500 mil. SKK Slovenia 8 (1994) 12.5 1994 ECU 3.1 mil.; 680 mil. SIT a. Expressed also in the ECU equivalent of local currency (using the October 2001 market exchange rate) if the requirement is in local currency. Source: World Bank. RESTRUCTURING ENTERPRISES Soft loans from banks are only one source of soft budgets. Even when banks are restructured and privatized and are no longer extending credit to unviable enterprises, the latter could still remain afloat through arrears to the state and other enterprises, which has implications for the fiscal budget. Arrears on tax payments have in fact taken the place of soft bank loans to become the most important source of soft budget constraints in some CEEC countries, as direct budgetary transfers have declined. In the Czech Republic, the stock of tax, social security, and health contribution arrears amounted to about 6.5 percent of GDP at end-1999.23 In Poland, tax arrears have become the single most important source of soft budget constraints for enterprises. It is X The explanation of the "extensiveness" and "effectiveness" of financial regulations can be found in European Bank for Reconstruction and Development 2001. Transition Report,2001, p.32. 23 European Bank for Reconstruction and Development 2001. Transihton Report,2001, p. vii. - 22 - Liquidating the Legacy of the Past estimated that in 1999 arrears on tax obligations and social contributions incurred by the coal- mining sector accounted for 0.2 percent of GDP. In Romania, state-owned enterprises (Petrom, Sidex, Termoelectrica, Tractorul, and the coal mining enterprises) have run up the largest arrears on taxes and social security contributions. Role and Status of Privatization One important route through which budget constraints could be hardened is privatization. A recent study reviewing 125 empirical studies on enterprise restructuring in the transition countries found that enterprise privatization is positively associated with enterprise restructuring (with the latter being measured in terms of either improvements in enterprise productivity or increases in the rate of production).24 The mode of privatization also matters. The same study found that the most effective owners were foreign investors. Foreign investors produce 10 times as much restructuring as diffuse owners, which were the least effective, and in fact were statistically indistinguishable from traditional state ownership. These empirical findings are consistent with the evidence in the CEECs. The countries that have involved foreigners to the greatest extent in the privatization process (and that are also among the top recipients of foreign direct investment) (that is, Hungary and Estonia) have also been among the best in economic reforms and performance. Most of the CEECs made rapid progress in the early years of transition in privatizing small firms, which were predominantly in the trade and service sector. By 2001, according to the EBRD, almost all the CEECs had reached industrialized market economy standards in terms of small-scale privatization, with only Bulgaria and Romania lagging slightly behind (table 2.6). Progress has been slower with respect to large-scale privatization, in part because of the economic and social constraints related to restructuring large-scale enterprises that are not commercially viable.25 By 2001, fewer than 50 percent of the large-scale enterprises had been privatized in five out of ten CEECs. Progress has been even slower with respect to corporate governance and enterprise restructuring; that is, in terms of the extent of financial discipline on enterprises through hard budget constraints (tight credit and subsidy policies) and/or an effective bankruptcy regimne. 24 See Djankov, Simeon, and Peter Murrell (2000). The Determinants of Enterpnse Restructurnng in Transition. Washington, D.C.: World Bank. The study reviewed 125 such empirical studies, and concluded that the move from state to private ownership of large state enterprises has resulted in greater amounts of restructuring. 25 European Bank for Reconstruction and Development. 2001. Transition Report,2001. - 23 - Expenditure Policies Toward EU Accession Table 2.6 Enterprise Privatization and Restructuring, 2001 Small-scale Large-scale Govemance and enterprise privatization privatization restructuring Bulgaria 4- 4- 2+ Czech Republic 4+ 4 3+ Estonia 4+ 4 3+ Hungary 4+ 4 3+ Latvia 4+ 3 3- Lithuania 4+ 3+ 3- Poland 4+ 3+ 3+ Romania 4- 3+ 2 Slovak Republic 4+ 4 3 Slovenia 4+ 3 3- Note: The EBRD indices range from 1 to 4+, with 4+ bemg the highest and representing industrialized market economy standards. Source: European Bank for Reconstruction and Development. 2001. Transition Report, 2001. Targeted Support to Large-Scale Restructuring Governments may need to incur fiscal costs in the short term to address the social consequences of restructuring large-scale enterprises in order to reap the benefits of lower fiscal costs over the long run that such restructuring would bring. Restructuring Coal Mines The Polish coal mines provide an example of what such a task entails (box 2.1). The cost of the four year program (1998-2002) amounts to around US$2 billion (1.3 percent of GDP in 1998 prices), consisting of about US$1.5 billion (around 1 percent of GDP) for severance costs and US$0.5 billion (0.3 percent of GDP) for physical mine dosure. Even when there costs are taken into account, the coal sector restructuring program has progressively reduced the drain on the Polish government budget as a hard budget is being applied. - 24 - Liquidating the Legacy of the Past Box 2.1 Restructuring the Polish Coal Mines The Polish coal mining sector-accounting for about 4 percent of GDP in Poland-has been experiencmg increasing difficulties over the last two decades. Between the early 1980s and 1998 the sales of Polish coal were nearly halved owmg to ongoing changes in energy efficiency and substitution (gas for coal) on the domestic front and a depressed global coal market and declme in export competitiveness on the external front. The mining companies failed to adjust to the fall in demand and, saddled with obsolete equipment, made mcreasmg losses and had to be kept afloat through arrears on various government obligations, including taxes, social security, and environmental fees and fines. By 1998, the cumulative total of such liabilities had reached around 3 percent of GDP. The government introduced a hard coal sector restructuring program in 1998 aimed at reversing the losses of the coal minng companies and preparing them for privatization while at the same time putting in place a safety net for those coal miners who would become unemployed as a result of the restructuring. Specifically, the program involves four elements: closure of mines, rescheduling or forgiveness of liabilities to the government and government institutions, the use of incentive-based management contracts (which are fixed term and include dismissal of management in the case of poor performance), and financial support to encourage voluntary retirement and active labor market programs to assist in the redeployment of coal workers. By end-2000, the program had resulted in employment cuts of 36 percent and a reduction m production capacity of 20 percent. Employment downsizing and labor cost reductions exceeded expectations. More than half of the workers that had left the industry have reached retirement age, while most of the rest have been reabsorbed into the local labor market. However, the financial performance of the coal mining companies continued to be poor because of poor management and the companies were not able to pay all their taxes and fees to government entities in 1999 and 2000. In response, the government introduced revisions to the program in 1999 to provide for additional capacity and employment reduction, to strengthen company management, and to delay the meeting of profitability expectations by one year to end-2001. The industry has started to achieve a profit on coal sales in 2001 although it is still making losses overall due to high financial charges and non-coal operating costs. A recent evaluation of the program indicates that extensions of the revised program will be needed to prepare the coal mines for privahzation. These extensions include further employment restructuring and mine closure as well as additional financial support to encourage workers to leave the sector. Sources: World Bank (1999) Report and Recommendation of the President on a proposed Hard Coal Sector Adjustment Loan to the Republic of Poland, May 18; World Bank (2001(d) Implementation Completion Report on a Hard Coal Sector Adjustment Loan to the Republic of Poland, June 2Z and World Bank (2001(e) Report and Recommendation of the President on a proposed Second Hard Coal Sector Adjustment Loan to the Republic of Poland, June 26. Some of the key lessons from the Polish experience, as well as from the experience of other countries that have undertaken similar large-scale industrial restructurings, are worth noting. * Such a task is extremely difficult and takes many years to accomplish. This means that a high level of commitment by the government is necessary. The Polish coal mining restructuring program that is currently being implemented is the fourth such program that has been put forward over the last decade, and the first one that has actually been implemented. * It is important to involve the affected workers in designing the program. One of the key elements in the success of the Polish program is the intensive dialogue between the government and the trade unions in the design of this program. * There is a need for adequate measures, including the funding of such measures, to address the social consequences of restructuring. In Poland, fast-acting and - 25 - Expenditure Policies Toward EU Accession targeted income support with demand-driven labor redeployment programs are integral elements of the task. * Given the dynamic nature of enterprise restructuring on such a scale, the government needs to be flexible and to be ready to adjust the program when needed. This was the case in Poland, where the program has already been adjusted once (box 2.1), and where the new government that came to power in 2001 is looking into the need for further adjustments with a view to privatizing the mines in a few years. Finally, macroeconomic stability and sound economic performance would help, as it is easier for redundant workers to find alternative employment in a growing economy. Restructuring Railways Railways present another important fiscal burden on the budget. Since the transition railways in the CEECs have experienced declines in labor productivity and rising financial difficulties. The chief underlying reason is that the transition has caused a shift from transport- intensive heavy industries toward a more service-based economy, while at the same time more competitive modes of transport have emerged (trucks for freight, and cars and airlines for passengers). These two factors hav~e led to a decline in the demand for rail transport that, in the absence of a corresponding reduction in the railway labor force, has resulted in declines in labor productivity. Table 2.7 shows that, for the eight CEECs in the table, in all but one country (Estonia) rail freight traffic has fallen faster than GDP, while rail passenger traffic has fallen faster than GDP in all of them. In addition, the politically driven underpricing of passenger transport, together with the fall in the demand for freight transport, has meant that the railways can no longer cross-subsidize passenger transport with revenues from freight. Table 2.7 Rail Transport: Volumes and Labor Productivity, 1999 (1988 = 100) GDP Toni/km Passenger/km Labor productivity index index index index Bulgaria 68 30 47 62 Czech Repubhc 95 38 51 70 Estonia 83 97 16 195 Hungary 100 32 59 96 Poland 122 46 50 83 Romarnia 71 23 36 51 Slovak Republhc 103 38 51 70 Slovenia 106 64 40 103 Source: GDP index from European Bank for Reconstruction and Development. Transitton Report Update, May, 2000. Taken together, these factors have resulted in increasing financial difficulties for the railways, which in turn have translated into a sizable burden on the government budget. Table 2.8 provides estimates of the operating deficits of selected CEEC railways in 1997. The cost to the government arising from these deficits ranged from 0.04 percent of GDP in Bulgaria to around 0.8 percent in Hungary. It should be noted that these costs exclude investment costs. However, the latter can add substantially to the total costs to the government. For example, a - 26 - Liquidating the Legacy of the Past separate estimate of the cost of railways to the government in the Czech Republic indicates that it amounted to 1.6 percent of GDP in 1998, inclusive of investment costs. Despite the sizable drain on the budget of these costs, the railways are in poor shape because of inadequate maintenance. Table 2.8 Operating Deficits of Selected CEEC Railways, 1997 (percent of GDP) Bulgaria 0.04 Czech Republic 0.62 Estonia 0.19 Hungary 0.77 Poland 0.46 Romania 0.52 Slovak Republic 0.62 Slovenia 0.69 Source: International Railways Union. Fundamental reforms in the way the CEEC railways are organized and operated is needed to address the problems of poor finances and low productivity, and to halt their continued drain on the budget. Fortunately, the organizational and financial changes required are the same as those mandated by the adoption of the acquis communautaire. Key steps include separation of the infrastructure costs (and perhaps the actual infrastructure institution) from the operating functions of the railways, separation of the accounts for passenger services from those for freight, and separation of the accounts for social services from those for commercial services as well as prohibition of subsidies to commercial services. In addition, CEEC governments will need to introduce competition into railways. They will need to establish regulatory regimes to ensure nondiscriminatory access to the infrastructure for all licensed operators, and to develop systems of access prices that generate adequate earnings for the infrastructure operator while returning appropriate signals to the operators. Such large-scale restructuring is costly. For coal, however, there will be much larger benefits of lower fiscal costs over the long run. By far the largest need for financial support would be assistance for the redundancy reduction of labor. For example, just the first stage of railway restructuring in Poland (box 2.2) required a 25 percent reduction in the labor force (about 40,000 jobs) at a cost of around US$200 million (0.1 percent of GDP). Restructuring the entire Polish railway system may eventually require a 50 percent reduction in employment, totaling about 300,000 jobs, and US$1.5 billion (0.9 percent of GDP). In addition, there will be the costs of rehabilitating and upgrading facilities, although such investments should only be made when the future traffic flows are known with reasonable likelihood, and the railways themselves should be able to absorb such costs. - 27 - Expenditure Policies Toward EU Accession Box 2.2 Restructuring Polish Railways The collapse of the planned economy had a wrenching impact throughout the Polish industrial sector, especially in coal, steel, and transport, and in transport especially in the rail subsector. Compared with 1988, Polish railways (PKP) carried in 1999 only 46 percent of the ton/km and 50 percent of the passenger/km. This is somewhat similar to other Central European countries. This sharp decline in traffic led to financial losses exceeding US$1 million per day. Driven by the downward spiral of PKP output and the resulting deterioration in its physical and financial condition (the 1999 PKP deficit nearly reached 2 percent of the country's GDP), the Polish Parliament on September 8, 2000, passed a law on the commercialization, restructuring, and partial privatization of the company. The reform aimed to encourage the provision of rail transport services that are more responsive to the needs of the market economy, to reduce the burden on the state imposed by PKPs heavy losses, and to help prepare the transport system for Poland's entry into the European Union. Within this reform the PKP plans to restructure its labor force by laying off some 37,000 workers (out of 205,000 in mid 1999) to cut its costs, facilitate the transition to a new sectoral organization, and make subsidiary units more viable for privatization. By end-2001, the program through its well-developed social program and active labor market policies resulted in employment cuts of 20 percent, and subsequent improvement of the employment structure and an increase in productivity. The commercialization and privatization process is now under way. Estonia and Poland are the most advanced among the EU accession countries. Lessons leamed made it clear that it is important to have the following characteristics: * Introduction of the reform programs in the early stages of economy transition * Social dialogue on each level and at each stage of the implementation process * Not only high management, but also middle management, dedication to the restructuring program * Active labor and region-oriented policies rather than social protection policies * Strong limitations on wage increases * Privatization as a main goal of the reform Finally, with cross-subsidization banned, governments will remain confronted with the direct cost of whatever subsidy they want to maintain for passenger services. One option to contain these remaining costs is being pursued by the Slovak Republic, specifically through decentralization of the responsibility for funding Public Service Obligations for passenger services to the regions that benefit from the services. Looking beyond the acquis communautaire, a number of EU countries are currently exploring broader roles for the private sector in their railways, especially for the competitive provision of local passenger services but also for privatized freight operators and even private providers of infrastructure (as in the United Kingdom). Among the CEECs, Estonia has been the most aggressive advocate of private railways, having concessioned its light density regional passenger operator and the infrastructure and freight operations on the main line railway connecting Estonia with Russia. Poland has publicly announced plans to concession two suburban passenger operators and is proceeding with concessioning a broad gauge iron ore line and the privatization of the main freight carrier. The experience in Estonia and Poland suggests that concessioning and privatization require significant investment in system analyses, the collection of information for potential investors, and the development of appropriate regulatory frameworks and enforcement. - 28 - Liquidating the Legacy of the Past Reforning Utilities A final major source of soft budget constraints in some CEECs is arrears to other enterprises, in particular to state-owned public utility companies. Such arrears also eventually find their way back to the budget, since it is difficult for governments to cut off support to public utilities given their economic and social importance. The energy sector in Romania is one example of efforts among the CEECs to address the issue of losses in the utility sector (Box 2.3). Here, again, the adoption of the acquis communautaire is of help, because it encourages such measures as the unbundling of utilities, the restructuring of tariffs and the elimination of cross subsidies, and open access to networks. Lithuania's power sector, for example, completed this process in January 2002. Box 2.3 Restructuring the Energy Sector in Romania In 2001, the total losses of the main thermal power producer Termoelectrica (which supplies more than 50 percent of consumed electricity and about 40 percent of heat to industrial users and district heating companies) are expected to reach about 2.2 percent of GDP, while losses in the natural gas sector are expected to reach 4.5 percent of GDP. In both cases the losses have arisen from the failure to adjust energy tariffs and poor collection and payments discipline. They have resulted in contingent liabilities for the government as the latter has extended guarantees for foreign borrowing of these enterprises. To stem the losses, the Romanian government is adjusting heating, electricity, and gas prices, which would also help attract private investors into those sectors. To mitigate the impact of the price adjustments on poorer households, the government has introduced a heating subsidy, and will also have in place a means-tested mmnimum income guarantee scheme. The government is also taking steps to substantially raise the collection rates through the establishment of escrow accounts to ensure that a fixed percentage of the final customer payments, as well as subsidy payments by the government, is immediately transferred to the main thermal power suppfer and the two gas companies, circumventing the local heat-distribution companies that have poor payment records. All these measures together should increase pressure for the privatization or liquidation of state-owned enterprises, which are the main beneficiaries of these implicit subsidies from the state. The Romanian government intends to privatize gas and electricity as the long-term solution to the problems of losses in the sector. In shifting its role from owner to regulator, the government is also planning legal changes to enable the gas and electricity regulatory agencies to function with independence and autonomy. Souirce: World Bank. Furthermore, going again beyond the acquis communautaire, many CEECs are actively mobilizing the private sector to participate in the provision of infrastructure services. Telecommunications, gas, and power utilities have been or will soon be privatized in countries such as Lithuania, the Czech Republic, Hungary, and the Slovak Republic. It is too early to assess the results of these recent privatizations, but if the experience of other countries around the world is any guide, private sector participation in the provision of infrastructure services should boost efficiency and ease the strain on public finances while increasing the coverage of such services.26 26 World Bank. 2001. World Development Report: Building Institutionsfor Markets. New York: Oxford University Press.. Private sector participation in infrastructure needs to be complemented by independent regulatory agencies that follow transparent procedures and are subject to oversight by a strong and independent judiciary. To improve the access to infrastructure services for the poor, governments can set investment targets at the time of privatization, allow for more flexibility in the price/quality combination in regulating infrastructure services, liberalize the entry of service providers, and provide subsidies in a targeted fashion. - 29 - Expenditure Policies Toward EU Accession CONCLUSIONS Liquidating the legacy of a profligate past can be an expensive proposition. Even when capitalized and amortized over time the fallout of bank restructuring has often hovered around an annual 1-2 percent of GDP. The costs associated with downsizing mines or railways can be similarly daunting, and the conclusions drawn are significant. * It is better to recognize problems early, to see them for what they are, and to tackle them. One reason why the cost of bank restructuring was so much lower in Estonia than, for example, in the Slovak Republic, is that Estonia nipped banking distress in the bud, while the Slovak Republic let the situation fester for years. In another area, one reason why Poland was able to downsize its coal sector is that it recognized the sector's problem as a social one and dealt with it as such. The same would apply in many cases of railway restructuring. * While the costs may be high, it pays to face them up front. In about every CEEC, cleaning up bank portfolios proved key to the resumption of growth. Similarly, restructuring infrastructural enterprises is essential if they are to modernize and provide the services that economies require to compete within Europe. More often than not, financing instruments are available from domestic or foreign sources to stretch the related charges over longer periods. * No blueprint has as yet emerged of the best way to deal with bad debt recovery. Recovering debt is important, perhaps even more for the principle (avoiding moral hazard) than for the amounts involved (nowhere in the region have actual recoveries exceeded 20 percent of the amounts due). In the circumstances, an eclectic mix of centralized and decentralized approaches would seem to recommend itself. * The key role played by foreign investment. Foreign banks in particular have played a central role in breaking up the "old boys' club" that used to bind banks to their borrowers and that proved fatal to the banks. Perhaps the now massive presence of foreign banks (they control more than two-thirds of the bank credit in most CEECs) will help the banking industry grow on a sound basis. Indeed, now that the stresses are subsiding, attention needs to shift from resolution to prevention. Many future crises and their attendant fiscal costs will be avoided if CEECs continue to enhance the implementation of their regulatory apparatus and the effectiveness of their debt-resolution procedures. - 30 - 3. SUCCESSFULLY INTEGRATING WITH THE EUROPEAN UNION The future, as much as the past, is a source of expenditure challenges of its own, and the future for the CEECs is to make a success of their ongoing integration into the European economy. To achieve this, these countries will need to upgrade the skills of their labor forces, the levels of their environmental standards, and the quality and extent of their transport networks. The nature of the fiscal challenges involved differs from one case to another. For instance, there is a clear need for countries to invest in environment and transport. As seen in chapter 1 some countries are already investing heavily. However, the question is to invest judiciously; that is, in a manner that maximizes the trade off between costs and benefits. In the case of skills and education, in contrast, the issue is a more difficult one of redirecting expenditures from yesterday's priorities to those of today and tomorrow. While the benefits of expanding parts of the education system may be evident, the problems associated with shrinking other parts may often be politically crippling. We will examine education, environment, and transport in turn. It should be kept in mind that, although expenditure policies will generally be discussed in pecuniary terms, their benefits extend well beyond narrowly defined financial gains. Indeed, for the man on the street, their impact may be more immediately felt in terms of broader access to knowledge, increased mobility, and cleaner water and air. UPGRADING SKILLS Human capital has been and will remain a key asset for securing the benefits of European. and global integration. Together with supportive macroeconomic and financial policies and infrastructure investments, education will continue to play a key role in supporting the transformation of the CEECs from low-income, resource-based economies to high-income, knowledge-based economies. In the last decade the demand for education was affected by two major changes: a collapse in the school age population and a sea change in the demand for skills. The supply side, however, was slow to react to these new conditions. As a result, teacher/student ratios in basic education, already low by international standards, fell further, while education systems struggled to keep up quality. The sluggishness of the supply response has much to do with built-in rigidities in financial allocation formulas, as well as in the politico- institutional frameworks that govern the sector. How well the CEECs reorient their education systems to the needs of the new regional and global economy will have major implications for their competitiveness in the future. - 31 - Expenditure Policies Toward EU Accession Demand Side Shocks The CEECs have already attained some of the highest levels of educational coverage at the start of transition. In most of the CEECs Education was compulsory for 11 years in most of the CEECs. Schools and universities were relatively well provisioned and maintained. Higher education programs in science and engineering were well developed and accounted for the bulk of higher education enrollments. Access to higher education was strictly controlled, which helped ensure that the quality of higher education programs was maintained at desired levels. The transition brought major disruptions in the demand for education. The skills that the economy required began to change rapidly, while the school-age cohorts started to contract at rates formerly seen only in cases of war, famine, or pestilence. Skills Match Transition and global integration brought about a demand for new and higher levels of skills. New technologies and new markets affected the human capital required from the labor force, making the skills of many manufacturing and mining workers irrelevant to new needs and creating the demand for other skills, particularly in the service sector. Furthermore, the decompression of wages that accompanied the transition boosted the implied returns on education, and hence the demand for it. This changing demand for skills called for fundamental changes in the content of education programs, in particular in making education programs more flexible, more student centered, and more focused on problem solving and the application of concepts rather than on the memorization of facts. Shrinkage of School-Age Population In parallel, however, the CEECs experienced a dramatic contraction in their school-age population. By the beginning of the transition, fertility had already fallen below replacement rates in most of the CEECs, and it continued to fall throughout the 1990s. As a result, the size of the school-age cohort contracted sharply (table 3.1) and will continue to shrink significantly for at least another decade. By the year 2015, the school-age cohort in the CEECs will be from one- third to one-half less than it was in 1990, and about 25 percent less than it was in 2000. Table 3.1 School-Age Population, 1990 -2015 Size of 0-14 year cohort (thousands) Change (percent) 2015 1990 2000 (projected) 1990-2015 Bulgaria 1,781 1,279 916 -48.6 Czech Republic 2,223 1,695 1,273 -42.7 Estonia 349 254 184 -47.3 Hungary 2,098 1,705 1,303 -37.9 Latvia 573 418 280 -51.1 Lithuania 841 723 546 -35.1 Poland 9,574 7,462- 6,185 -35.4 Romania 5,468 4,112 3,154 -42.3 Slovak Republic 1,351 1,070 824 -39.0 Slovenia 381 318 243 -36.2 Sources: World Bank world development indicators and demographic databases. - 32 - Successfully Integrating with the European Union Enrollments In spite of the difficulties experienced by the accession countries during the 1990s, official data on school enrollments show improved coverage of preschool, primary, and, especially, higher education-expressed as a percentage of the relevant age group enrolled in school-in most of the accession countries (table 3.2). However, registered enrollments actually declined as a percentage of the age group in two countries at the preschool level (Lithuania and Slovakia), in four countries at the primary level (Bulgaria, the Czech Republic, Hungary, and Latvia), and in five countries at the secondary level (Bulgaria, the Czech Republic, Latvia, Lithuania, and, especially, Romania). Table 3.2 Enrollment Ratios through the Transition Preschool net Primary gross Secondary gross University gross enrollment ratio enrollment ratio enrollment ratio enrollment ratio (percent) (percent) (percent) (percent) 1990 1999 1990 1999 1990 1999 1990 1999 Bulgaria 65.5 66.4 98.6 95.1 77.0 75.6 26.2 34.7 Czech Repubhc 75.2 85.4 98.6 97.7 78.7 75.9 17.2 26.0 Estoma 67.4 73.5 94.9 97.5 57.0 71.9 34.4 45.0 Hungary 85.3 87.3 98.8 98.7 73.3 98.8 12.1 28.9 Latvia 44.8 61.0 94.9 92.3 70.2 68.5 20.5 46.5 Lithuania 55.9 51.6 92.5 95.5 70.0 64.8 26.5 39.2 Poland 47.1 49.9 97.5 98.3 89.3 99.5 17.0 42.8 Romania 53.3 66.2 92.5 98.5 89.9 70.2 9.2 23.4 Slovak Republic 72.0 69.5 98.1 107.5 78.2 80.0 14.3 22.5 Slovenia 55.7 70.2 95.3 97.4 n.a. 93.3 22.9 51.0 Note: Figures shown are gross enrollment ratios, which tend to overstate actual coverage because they include overage students in the numerator but not in the denominator. Source: UNICEF, 2001, A Decade of Transition: The MONEE Project. CEE/CIS/Baltics, Innocenti Research Center. Most of the registered decline in secondary enrollments occurred in vocational and technical education. In Romania, for example, the enrollment in secondary vocational and technical education declined from 78.4 percent of the age group in 1990 to 43.9 percent in 1999, while the enrollment in general secondary education increased from 11.5 percent of the age group to 26.3 percent in 1999.27 The decline in secondary enrollments reflects a perception that secondary education-especially secondary vocational education-no longer ensures employment for graduates. (Many- of the enterprises that had traditionally recruited secondary vocational students at the completion of their training closed or reduced their staffing, and no longer recruited graduating students.) It may also reflect the pressure for some students to enter the labor market in order to augment falling household income. In the poorest of the CEECs, the shifting of financing responsibilities to households for textbooks and other educational tools that were formerly provided free led to financial hardship for some households. This may have contributed to the declines in coverage that were observed for primary and secondary education in Bulgaria and for secondary education in Romania. 27 Data provided by Romanian Ministry of National Education. - 33 - Expenditure Policies Toward EU Accession School Attendance Although the official enrollment data show a generally improving coverage of education, household survey data often show that actual school attendance is well below registered enrollments and is declining, and reveal significant gaps in attendance in rural areas, in areas with ethnic minorities, and among the poor. In Bulgaria, for example, school attendance rates are lower for the rural population than for the urban population, especially for secondary education (where remoteness of schools is often a constraint); school attendance rates for the poor are much lower than for the non-poor at all levels of education; and Roma have much lower rates of school attendance than either ethnic Bulgarians or Bulgarian ethnic Turks for all levels of education (table 3.3). Table 3.3 Bulgaria: Rates of School Attendance by Level, 1995,1997, and 2001 Preschool education Primary education Secondary education 1995 1997 2001 1995 1997 2001 1995 1997 2001 Total population 44 14 22 87 88 90 47 55 46 Males 42 12 21 88 88 90 49 54 46 Females 46 15 24 85 88 89 45 56 46 Urban 46 13 24 88 90 92 52 63 53 Rural 40 14 20 83 84 84 31 32 22 Non-Poor 47 16 26 89 93 94 49 60 52 Poor 8 11 10 54 81 70 20 46 13 Bulgarians 44 15 26 90 93 94 55 66 56 Turks 53 10 19 88 93 90 10 30 34 Roma 25 5 16 55 58 71 3 5 6 Source: Bulgaria Integrated Household Survey, 1995,1997,2001 data. Policy Responses The situation just described suggests that the following actions should be pursued: * Education at the secondary level should provide more generic skills for a few broad families of occupational specializations rather than highly specific skills for a large number of narrow occupations. * Vocational education should place more emphasis on developing numeracy skills, problem-solving skills, communication skills (including foreign language proficiency), and teamwork skills, and should give less emphasis to job-specific skills. * Higher education should expand and become more flexible at entry and should also offer easier transfer opportunities across programs and faculties, and should also provide stronger performance incentives to students and faculty. * The legal and fiscal environment should change so as to encourage employers and local governments to develop lifelong learning programs to meet local (and global) skill needs. - 34 - Successfully Integrating with the European Union Education Financing The resources available to undertake this redirection have remained constrained. By and large, the levels of public expenditure on education remained broadly stable in 1995-2000. When they have changed, it has tended to be downward (table 3.4). Bulgaria, the Czech Republic, and Romania experienced particularly severe declines in real expenditures for education. Table 3.4 Public Expenditures on Education, 1990-2000 Expenditures on education Real expenditures on educationa (Percent of GDP) (1995 = 100) 1995 2000 2000 Bulgaria 4.1 4.1 76.6 Czech Republic 5.2 4.2 81.1 Estonia 7.5 6.5 119.0 Hungary 5.2 4.8 105.4 Latvia 6.6 6.5 134.2 Lithuama 5.4 5.8 101.5 Poland 5.2 5.9 136.5 Romania 3.4 3.1 83.3 Slovak Republic n.a 3.9 90.2 Slovenia 5.5 5.4 118.5 Note: Expenditure figures refer to consolidated general budget. a. Deflated by the implicit price of government consumption. Source: World Bank database. As will be described, decentralization has had a crucial-often-negative-effect on the resources available for education at the local level. Conversely, resource availability has been improved through various actions taken to diversify the sources of financing for education programs, including requiring parents to purchase textbooks and other educational materials formerly provided free by schools, expanding private education, instituting student fees and other charges for cost recovery in specialized secondary education and higher education, and allowing schools to raise and retain funds through such actions as the rental or sale of unneeded facilities and the provision of extracurricular courses. Resource Utilization Within a limited resource envelope, however, the success of the strategy for redirecting education that has been outlined was predicated on the capacity of governments to shift resources from lower to higher levels of education, including the downsizing of staff and facilities in many primary and secondary education programs. In practice, changing allocations in this manner has proved problematic. Indeed, the response of education managers throughout the CEECs has more often than not focused on maintaining the jobs of teachers and other educational staff. Already low student/teacher ratios fell (table 3.5), and therefore teacher workloads fell significantly at all levels of the education system, signaling major room for improved efficiency. - 35 - Expenditure Policies Toward EU Accession Table 3.5 Changes in Student/Teacher Ratios StudenVtteacher ratio in primary education 1990 1997 Bulgaria 14.8 13.9 Czech Republic 19.6 14.5 Estonia 10.5 11.7 Hungary 12.5 12.2 Latvia -n.a. 12.0 Lithuania 12.0 11.3 Poland 16.7 15.4 Romania 16.7 14.8 Slovak Republic 19.4 17.1 Slovenia 15.4 13.5 OECD average 17.1 Japan 21.4 South Korea 31.0 United Kingdom 22.0 New Zealand 24.7 Sources: Population growth from World Development Indicators. Accession country student/teacher ratios from UNICEF-ICDC database, as provided in Berryman, Sue E. 2000, Hidden Challenges to Education Systems in Transition Economies. Washington, D.C. World Bank; OECD student/teacher ratios from OECD, 2000, Education at a Glance. Paris. With the shrinkage of school-age cohorts during the 1990s and the focus on maintaining existing teaching positions, class sizes and student/teacher ratios-already below OECD levels at the start of the decade -dropped to very low levels, implying major inefficiencies, by the end of the decade (table 3.5). The dispersed rural population complicates the task of school rationalization. In Lithuania, for example, 13 percent of rural comprehensive schools have an average of five students per class in grades 6 through 9, 23 percent have seven students per class, and 31 percent have ten students per class.28 Country norms on minimum teaching hours in the CEECs are also low by OECD standards. For example, teaching hours in primary schooling average 583 hours per year in Hungary and 724 hours in the Czech Republic, versus 958 hours in the United States and 788 hours in the OECD as a whole.29 In this context, some of the increases in the "volume" of education observed in table 3.3 may well reflect more "input" developments (related to levels of staffing) than "output" developments (reflecting how much education is actually being delivered). In the Slovak Republic, for example, the number of staff employed in the education sector increased-in the face of declining overall enrollment-by 5 percent in primary education, 20 percent in secondary vocational education, and 32 percent in general secondary education over the period 1990-99. One of two outcomes (or a combination of both) typically ensued: (a) Falling real teachers' salaries leading to serious problems with teacher morale and motivation, as well as to 28 Economic Research Centre. 1999. "Effectiveness and Efficiency of Public Expenditures in the Education Sector, Summary." Vilnius, May 1999, cited m OECD 2002. Lithuania, Education and Skills: Reuiew of National Policies for Education. Paris. 29 OECD. 2000. Education at a Glance. Paris. - 36 - Successfully Integrating with the European Union growing problems of corruption in some of the CEECs (with students paying for good grades or entrance into the most desirable programs) and (b) outlays for salaries and benefits rising as a share of recurrent education expenditures and in the process displacing vital expenditures to maintain and update teaching, learning materials and school infrastructure. Impact Partly as a result of these conditions, both the quality of and access to education, while still high, are beginning to show signs of eroding. Changes in Quality For the CEECs, the most inclusive source of internationally comparable data on what students learn is the TIMSS international science and math assessment.30 Eight CEECs -Latvia, Lithuania, Hungary, Bulgaria, Romania, the Slovak Republic, Slovenia, and the Czech Republic-participated in the 1995 and 1999 surveys. As shown in table 3.6, five of the CEECs (the Czech Republic, Slovakia, Slovenia, Hungary, and Bulgaria) had mean math and science scores above the international average in 1995. Latvia, Romania, and Lithuania all scored well below the average. Table 3.6 TIMSS Eighth Grade Student Assessment Results for Science and Math for Eight CEECs, 1995 and 1999 Mathematics Science 1995 mean score 1999 mean score 1995 mean score 1999 mean score Czech Republic 546 (4.5) 520 (4.2) 555 (4.5) 539 (4.2) Slovak Republic 534 (3.1) 534 (4.0) 532 (3.3) 535 (3.3) Sloverua 531 (2.8) 530 (2.8) 541 (2.8) 533 (3.2) Hungary 527 (3.2) 532(3.7) 537 (3.1) 552 (3.7) Bulgaria 527 (5.8) 511 (5.8) 545 (5.2) 518 (5.4) International average . 519 (0.9) 521 (0.9) 518 (0.9) 521 (0.9) Latvia 488 (3.6) 505 (3.4) 476 (3.3) 503 (4.8) Romama 474 (4.6) 472(5.8) 471 (5.1) 472(5.8) Lithuania 472(4.1) 482 (4.3) 464 (4.0) 488 (4.1) Note: Standard errors are in parentheses. Sources: International Association for the Evaluation of International Achievement, 2000, TIMSS. 1999: International Mathematics Report, December; International Association for the Evaluation of International Achievement, 2000, TIMSS 1999: International Scwnce Report, December. The Czech Republic's average math and science scores dropped sharply in 1995, but remained well above the international average, and Slovenia's fell slightly. Bulgaria's average scores in both tests fell very sharply, in both cases falling below the international average. Latvia and Lithuania's average math and science scores improved significantly in 1999, but not enough to take them above the international average. Only Hungary initially succeeded in taking its scores further above the international average (more recent test assessments not reported here show similar signs of quality deterioration). 30 The survey was carried out for a nationally representattve sample of eighth-grade students in 24 countries m 1995 and 39 countries in 1999. - 37 - Expenditure Policies Toward EU Accession Two important messages emerge from these TIMSS assessment results. The first is that is that students in the CEECs perform quite well in relation to the international average in this specific application. This result largely reflects the legacy of discipline and pedagogy in the inherited education programs. A second message is that student scores on the TIMSS assessment have deteriorated in several of the CEEC countries, and resource deprivation appears to have contributed to that decline. The drop was most pronounced in Bulgaria and the Czech Republic, where real public expenditures on education at the end of the decade were lowest in relation to the level of education expenditures in 1990 (table 3.4). Another international assessment of what students learn is the Program for International Student Assessment (PISA),31 carried out by the OECD. Whereas the TIMSS assessment tests students' mastery of the formal curriculum, and the test questions follow the material as it is typically presented in class, the PISA test instrument deliberately aims to assess students' mastery of higher-order skills such as synthesizing knowledge across disciplinary boundaries, integrating uncertainty into analysis, monitoring their own learning progress, and knowing where to access relevant information. These are exactly the skills that are needed for most of the fastest growing jobs in the global economy, as revealed by the experience of the OECD countries and the CEECs themselves.32 As summarized in table 3.7, the CEECs perform much less impressively under the PISA assessment than they do under the more limited TIMSS assessment. The four CEEC countries that participated in the PISA assessment have scores well below the OECD average, and not much above the lowest two performers on the PISA assessment, Mexico and Brazil. One message emerging from these PISA results is that education systems in the CEECs cannot rest on their laurels, but must do a much better job of developing higher-order skills of synthesis and application. This will require a fundamental change in curriculum and teaching methods, involving a more student-centered, inquiry-based form of pedagogy, and more reliance on sources of information other than the textbook and the teacher's presentation of the approved curriculum. Table 3.7 PISA Student Assessment Results for Literacy for 15-Year Olds for Four CEECs, 2000 Total variation Of which inter- Mean score in student resultsa school variationa OECD average 500 100.0 36.0 Czech Republhc 492 97.2 51.9 Hungary 480 106.0 71.2 Poland 479 105.9 67.0 Latvia 458 112.6 35.1 Mexico 422 80.3 42.9 Brazil 396 82.9 35.8 a. Expressed as a percentage of the average variation in student performance in OECD countries. Source: OECD. 2001. Knowledge and Skills for Life: First Results from PISA 2000. Paris. There are also indications of increased between-school variation in education quality in the CEECs. The TIMSS study does not report between-school variation, but the PISA 31 This study was carried out by the OECD for a sample of 15 year olds in 31 countries (including Latvia, the Czech Republic, Poland, and Hungary) in 2000. OECD, 2001, Knowledge and Slls for Life: First Results from PISA 2000, Paris. 32 OECD and Statistics Canada, 2000, Literacy in the Information Age, Paris and Ottawa. - 38 - Successfully Integrating with the European Union assessment does. As shown in table 3.7, there is a greater dispersion of student assessment results for three of the four CEECs than for the OECD average. And in three CEECs-not the same three-much more of the variation in student assessment results is explained by between- school differences than is the case for the OECD average. These greater between-school differences in Hungary, Latvia, and the Czech Republic may reflect the impact of decentralization policies that reduced the uniformity in resource provision and content across schools that had existed prior to the transition.33 None of the internationally comparable student assessments covers higher education. But the quality of higher education has unquestionably suffered in many specializations during the 1990s as restrictions on access to higher education were relaxed in response to an enormous public demand. In all of the CEECs, enrollments in public institutions have grown much faster than budget allocations. As at other levels of education, priority has been given to maintaining teaching positions. Limited budget resources have been available to maintain essential teaching and learning inputs, including laboratory materials and library collections. The few resources that have been available for these purposes in higher education have tended to be used for the acquisition of (IT) hardware and software. Policy Agenda Falling secondary coverage in Bulgaria, the Czech Republic, Latvia, Lithuania, and Romania and shrinking expenditures for education in Lithuania, the Slovak Republic, and, especially, Bulgaria present serious risks of "deskilling" the labor force and undermining labor force productivity and flexibility in the future, as does the thinning of resources for higher education and research in all of the CEECs. Although there is a need for a more intensive use of staff in higher education, as at other levels of education, there is also a conspicuous need for more non-salary recurrent resources to finance educational materials, including laboratory and library materials. In general, resources need to be redirected and financing and management mechanisms changed to ensure the availability of quality education to the poorest households and localities, and to raise the quality of education at all levels. Some of the necessary resources should come from improved efficiency within education systems, particularly in the form of the more intensive use of teaching staff and the shedding of redundant teachers at all levels. The savings from these efficiency improvements should be used to increase quality. Increased teacher remuneration for a streamlined teaching corps can play a role in inducing improved motivation and teaching effectiveness. Financing Fornula Input-based financing formulas are the main reason why the rationalization of school networks has not made more progress in the transition countries. The current education financing formula in most of the accession countries finances educational inputs; that is, teachers, and sometimes textbooks, other educational materials, and in-service teacher training. 33 Unfortunately, there are no time series data that would allow us to track changes in these variables over time. But whatever their time trends and their causes, these sizable between-school differences in what students learn should be a cause for concern in countries that aspire to educate all students to international standards. - 39 - Expenditure Policies Toward EU Accession This financing model provides few efficiency incentives. It ensures financing for all current schools and teachers, as long as the class size and minimum teaching hour standards established by the education ministry are met. Furthermore, these statutory minimum class sizes and minimum teaching hours are well below OECD norms. In addition, education ministry inspectors can, and often do, provide exceptions that do not meet even these low standards. Not surprisingly, little school optimnization or system rationalization has been carried out in the CEECs under this model. The little system rationalization that has occurred-as in the Czech Republic34-has tended to take place only within schools and not across schools. Organized opposition can easily thwart local governments' attempts to carry out rationalization, as it did in the Czech Republic. A preferred method for financing education is capitation-based financing, which determines the amount of a local government's educational subsidy based on the number of students it is educating at each level, and differentiated to reflect different costs of different programs of education and possibly other sources of cost variation. This approach -used in the Czech Republic and Lithuania-is preferred for two reasons: first, because the basis of financing-enrolled students-is much doser to the educational objective than are inputs such as numbers of classrooms and teachers and, second, because it provides an incentive for providers to rearrange inputs in order to provide education more efficiently. The capitation approach is not perfect. It does not, by itself, provide safeguards to ensure education quality or teaching effectiveness. Neither does it reflect cost differences among different programs, place-specific cost factors, or cost differences arising from the special learning needs of students. Finally, it does not provide for improvements in curriculum, teaching materials, and teaching practices, all of which are needed in the CEECs. A composite financing formula (as in table 3.8) may provide for all of these needs.35 Among the CEECs the Czech Republic, the Slovak Republic, and Lithuania have moved farther in that direction. They finance primary and secondary education through capitation formulas with some of the elements shown in table 3.8 to reflect cost variations. The Czech Republic also uses a capitation formula to finance lifelong learning courses offered by universities. Romania and Bulgaria calculate per student costs, but do so ex post, and the financing formula remains therefore essentially input-based and carries the related drawbacks. 34 The Czech school rationalization program successfully consolidated 159 schools and reduced 4,000 jobs, but was suspended due to the opposition of teachers and local communities. 35 Ross, Kenneth, and Rosalind Levacic, eds. 1999. Needs-Based Resource Allocation in Education via Formula Funding of Schools. International Institute for Education Planning, UNESCO, Paris. Some of the most advanced applications of this approach are in the Anglo-Saxon countries: the United States, Canada, England, Wales, Australia, and New Zealand. - 40 - Successfully Integrating with the European Union Table 3.8 A Composite Formula for Education Finance Component Dimensions Indicators Basic per student allocation * Total enrollment, * Full-time equivalent enrollments by grade differentiated by and type of program grade and program School site needs * School size * Primary < 200 full-time equivalent * Secondary < 600 full-time equivalent * School remoteness * Kilometers to town of 50,000+ persons * Operations and * Interior area of school in square meters maintenance costs Student supplementary * Socioeconomic * Percent of students from households educational needs hardship receiving social assistance * Low educational * Number of students below 20th percentile achievement assessment results * Non-fluency in * Percent of students below cut-off score in national language national language test * Disabilities and * Number of students formaly assessed with special learning special learning needs needs Educational quality * Specialized * Full-time equivalent enroUed in specialized improvement curriculum program * Specialized school * Total full-time equivalent (if special curriculum school) Source: Adapted from Levaci, Rosalind, and Kenneth Ross. 1999. "Principles for Designing Needs-Based School Funding Formulae." In Kenneth Ross and Rosalind Levaeic, eds., Needs-Based Resource Allocation in Education via Formula Funding of Schools. International Institute for Education Planning, UNESCO, Paris. In the CEECs that have not yet adopted capitation-based financing, financing formulas for education at all levels should be based on the number of students rather than on inputs. Designing appropriate capitation formulas involves balancing delicate trade-offs: First, per student allocations should be differentiated to reflect the intrinsic differences in the cost of education delivery, such as the higher cost of technical specializations, and greater population dispersion in rural areas. Great care needs to be taken in designing this differentiation to reflect the unavoidable differences in the cost of education provision that are due to factors such as population dispersion and climate differences. It is essential that the financing formula reflect these differences in order not to impose further hardship on the districts that experience higher costs. Second, capitation formulas need to provide meaningful incentives to seek efficiency gains. If financing formulas simply mirror the current unit costs of different localities, then the resulting schedule of coefficients would merely legitimate an inefficient delivery model. The same considerations apply to the differentiation of costs for different programs of studies.36 36 In the Slovak Republic, for example, per student recurrent costs are 100 percent higher for upper secondary vocational education and sports education schools than for gymnasia. Per student costs in professional art schools are almost four times as high as in gymnasia. Whereas unit investment cost differences among specializations primarily reflect the technical requirements of the specialization, recurrent costs differ largely as a function of class size and teaching loads. Secondary art schools in the Slovak Republic typify the problem of unsustainably high costs that result from too small class sizes. The recurrent-cost financing formula for upper secondary and higher education should encourage these institutions to rationalize course offerings. This could take the form of moving toward a more affordable class sizes, or reconfiguring course offerings (for example, by providing art education as one of several options in comprehensive secondary schools rather than in free-standing art schools). In contrast, m the Czech Republic, per student allocations in upper secondary schooling range from about CZK 24,000 for gymnasia - 41 - Expenditure Policies Toward EU Accession A still more advanced approach to formula financing of education is to finance educational results rather than enrollments. Some of the charter school contracts in the UrLited States, for example, condition the payment to private education providers on the achievement of agreed educational targets in terms of learning achievement. The Czech model for subsidizing private education embodies the same approach. It finances a higher proportion of recurrent costs for schools that meet higher quality standards. This approach is likely to grow in use as the tools for assessing school performance improve. Decentralization and Education Finance Decentralization of education finance in the CEECs has typically involved financing teacher salaries from the state budget but devolving responsibility for school maintenance and the provision of educational materials (and sometimes even teacher training) to local governments. In principle, the decentralization of responsibilities for education finance and management to local governments offers the potential to make the management of education more efficient and the content of education more responsive to local needs. It could also encourage the mobilization of additional resources for education. In practice, however, inconsistencies between responsibilities and resources of local governments have often frustrated the potential benefits of decentralized education management in all of the CEECs. In all of the CEECs, local governments are meant to be accountable to the local community for managing basic education effectively and efficiently. But they sometimes lack the authority to do so because the bulk of financing-for teachers' salaries and benefits-remains centrally controlled in all the CEECs, and because ministries of education retain control over key decisions affecting education delivery. Ministries of education in the CEECs are responsible for curricula, recruitment, evaluation, training, and promotion of school principals and teachers, and for the establishment of norms governing minimum and maximum class size and teaching hours. These constraints make it impossible for local governments to carry out actions that will improve efficiency-such as school consolidation-unless the ministry of education agrees. In most of the CEECs, the ministry of education must approve any proposals for teacher dismissals, school closure, or school consolidation. In this setting, the actions taken by the CEECs to decentralize education finance and management and to diversify sources of financing have yielded little of the expected benefits, but have led to increased inequality in access and, particularly, in educational quality between rich communities and poor communities,37 and between rich households and poor households. and business academies to about CZK 29,000 for technical schools. This relatively narrow spread encourages more efficient de'avery of technical education. Because any additional costs would need to be financed from local sources, it also encourages local authorities to consider carefully whether technical education programs that cost more than this amount are providmg good value to the local community. 37 The most systematic differences in education quality in the CEECs are the differences in education quality between urban and rural schools. This gap reflects not only the poorer resource endowment of rural schools, but also the generally lower qualifications and experience of teachers in rural areas, the poverty of rural households, which makes it hard for them to afford education-related purchases, the lower educational status of rural parents, and the relative lack of educational stimuli in rural households. Cost differences exacerbate the resource differences between urban and rural areas. Costs of education are higher in rural areas than in urban areas because dispersion of - 42 - Successfully Integrating with the European Union In relying more heavily on financing from households and local communities, they have accentuated the pre-existing differences in quality of education and access to education between richer and poorer areas. Local governments vary widely in their capacity to mobilize resources from local taxes and other sources (including parental contributions). Thus, the shifting of financing responsibilities to local communities and to parents has meant that many households are not able to afford textbooks and other essential educational materials for their children, and many communities are not able to provide essential teaching and learning materials in their schools, or maintain their schools. There is a widespread perception in the CEECs that central budget financing of the salary costs of basic education is a temporary expedient which will eventually be replaced by local governments' assuming full responsibility for at least the recurrent costs of primary and secondary education. There is a major risk that such a move to complete reliance on local financing for basic education will lead to very negative consequences for poorer communities - including dosure of schools and the emergence of unacceptable quality differences in education. For these reasons, it appears preferable to maintain the central financing of teacher salaries and benefits and other essential recurrent costs of primary and secondary education, including textbooks and teacher training, rather than delegating these to local governments. The policy adopted in all of the CEECs essentially involves a weak version of this approach. This involves identifying a minimum set of essential educational inputs and financing them uniformly for all schools from central government resources. For primary and secondary education, recurrent costs (or at least teacher salaries and benefits) are financed by the state budget through tied central government grants to local governments that provide no discretion in how funds are spent. This formula varies in detail across the CEECs. In general, the financing formula provides for central budget financing of more educational inputs in the more prosperous CEECs than in the less prosperous CEECs. In the Czech Republic, the Slovak Republic, and Lithuania, for example, the central government budget finances not only teachers' salaries but also for school utilities, textbooks, in-service training for teachers, and teacher salaries and benefits for private schools that meet stipulated quality standards. This broad universal financing helps reduce, but does not eliminate, inequality in education.38 The risks of inequality in education are greater in those countries that finance only teachers' salaries from the central budget and leave other input to local governments or parents. In Bulgaria, for example, the poorest of the CEECs, parents are required to purchase textbooks. This clearly leads to differences in learning and may also affect school attendance. Specialized secondary education, higher education (apart from the retained income from student fees and other sources), and education for children with special needs is financed centrally in most of the CEECs, as are capital investments for schools. Universal central financing is more effective in the population leads to small class sizes, large transport costs, or both. Heating and utility costs also tend to be higher for rural schools. Because these intrinsic sources of higher unit costs of education are most prevalent in areas with the smallest revenue bases, they tend to reinforce the mequality m revenue capacity among localities. 38 For example, Lithuania provides a textbook allocation of LTL 20 (e 5.55) per student, but the actual cost of secondary textbooks is about LTL 150 (E 41.60) per year. OECD. 2002. Lithuania, Education and Skills: Revwew of National Policiesfor Education. Paris. - 43 - Expenditure Policies Toward EU Accession ensuring that schools continue to operate at the compulsory level despite differences in local capacity. But to be fully effective, it should be extended to cover textbooks and other educational materials that are essential to effective education. Most of the CEECs in principle rely upon local resources to finance essential educational inputs, and provide compensatory support through targeted subsidies to localities that could not otherwise afford to provide education of an acceptable quality. This targeting approach is also used in the United States, where the investment and recurrent costs of primary and secondary education are essentially locally financed, but state and federal subsidies are provided to communities with limited revenues and communities that face unusually high costs of education for children with disabilities or special learning needs. To some extent, equalization grants serve this purpose in the CEECs. But equalization grants are effective in ensuring that basic educational needs are met because they are untied, and thus can be used by local governments for other purposes. Tied grants would be more effective in ensuring that basic educational needs of poor communities are met. But this purpose could more logically be met by the tied (categorical) grants that are provided to local governments for specific education expenditures. The Problem of Excessively Small Local Governments The small size of local government jurisdictions is leading to serious problems in the management of education in the CEECs. Many local governments are too small to support schools that provide a reasonable critical mass and economies of scale, and too small to carry out effective system rationalization through school consolidation and school closure. Romania has 263 towns and municipalities and 2,688 communa that are eligible to receive central government transfers under the country's decentralization program. Hungary is the most fragmented of the CEECs, with 3,147 local governments, of which 2,443 operate schools. Some local governments in Hungary find it difficult to maintain primary schools, and some have closed their schools since the enactment of the Law on Local Self-Government.39 To address the problems of economies of scale of very small municipalities, the Czech Republic is in the process of establishing regional-level management for schools. This approach could help rationalize the allocation and use of education resources in the other CEECs with local governments that are too small to capture the efficiency gains of larger management units. Per student financing can also help address this problem, since the allocation can easily be transferred with students who attend schools in other jurisdictions. This is being done in Lithuania, for example, where a new financing formula provides for a "student basket" of inputs for each type of program, which can be transferred when students attend schools outside their own locality. 39 In principle, county governments are responsible for managing secondary schools in Hungary, but local governments may also run secondary schools. In 1993, the Hungarian Ministry of Education established eight Regional Educational Directorates. These Regional Directorates were expected to oversee educational programs at the regional level, but the Law on Local Self-Government gives such primacy to local governments that it proved impossible to give these regional bodies a substantive role. They were subsequently dissolved. - 44 - Successfully Integrating with the European Union Reorienting Education to Global Knowledge Needs Finally, the disappointing PISA assessment results for four of the more advanced CEECs indicate a problem regarding how students learn and what students learn. A fundamental change in the orientation of education is needed in order to support global competitiveness in the CEECs. A recent presentation to the OECD Governing Board described the nature of those changes.40 Effective education systems will need to teach the following: * Meta-cognitive abilities and skills; that is, thinking about how to think and learning how to learn. * The ability to integrate formal and informal learning, declarative knowledge (or knowing that) and procedural knowledge (or knowing how). * The ability to access, select, and evaluate knowledge in an information-rich world. * The ability to develop and apply diverse forms of intelligence. * The ability to learn and work effectively in teams. * The ability to create, transpose, and transpose knowledge. * The ability to cope with ambiguous situations, unpredictable problems, and unforeseen circumstances. * The ability to cope with multiple careers; that is, learning how to "redesign" oneself, locate oneself in a job market, and choose and fashion the relevant education and training for oneself. All the CEECs have stated an intention to reform their education systems to improve the quality and relevance of education programs. This is a long-term endeavor, requiring fundamental changes in all educational inputs, including developing improved curricula, textbooks, and teacher training programs; using student assessment as a tool of performance evaluation; developing streamlined vocational training prograrns, lifelong learning programs, and the accreditation of higher education institutions and programs; improving university entrance examinations; and providing more supportive role for the inspectorate. Although there has been some progress, it will be several years before these educational reforms bear fruit. Currticulum The financing and decentralization measures that have been implemented in the CEECs during the past decade are making it harder to bring about reforms in education programs. In Hungary, for example, curriculum formulation has been highly decentralized, which makes it difficult to assess or replicate the gains of innovation introduced at the local level.41 Slovenia 40 Presentation by Professor David Hargreaves, Cambridge University, to the OECD CERI Governing Board, March 24, 2000. 41 Halasz, Gabor. 2000. "System Regulation in School Education in Hungary." Report. National Institute of Public Education, Budapest - 45 - Expenditure Policies Toward ELU Accession alone among the transition countries is reforming its curriculum under a centralized process,42 which should facilitate the introduction of reforms. Central governments in the CEECs will need to play an active role in promoting the modernization of education systems. One approach is to provide a specific allocation in funding formulas, such as the allocation for research and development in the Czech financing formula for higher education institutions. This method is prone to waste, however, because it does not require the recipient institution to demonstrate the relevance of the action. A preferred approach is a competitive process that requires each institution to demonstrate the relevance of the proposed innovation. An example is Romania's central program for supporting curriculum innovation in higher education. This competitive approach is also the main method used for allocating federal financing for scientific research in the United States. Reforn of Vocational Education and Developing Lifelong Learning Vocational education in the-CEECs is in crisis; that is, struggling to cope with dwindling enrollments, increasingly irrelevant curricula, and a major loss of enterprises that are able to provide on-the-job training to students. The reform of vocational education needs to accelerate. The increasing share of secondary general education should be welcomed rather than resisted, since general education provides greater flexibility than specialized vocational education and since many of the skills provided in good programs of general education (such as foreign language skills, math, science, problem-solving, and teamwork) are exactly the skills most needed in the global economy. Governments should not attempt to predetermine the shares of students opting for general secondary education and vocational secondary education, but should provide parents and students with better information about the career implications of these choices, and should adjust the capacity of these programs to accommodate public preferences. At the same time, the existing programs of vocational education need to be restructured into fewer specializations for clusters of kindred occupations that draw upon the same basic competencies. Some of the CEECs have made a strong start in transforming their vocational education systems in this direction. Hungary's reform of vocational education provides a good example of how this might be carried out. It involves reducing the number of vocational specializations, increasing their science and mathematics content, developing new vocational curricula based on prevailing industry practice, and developing lifelong learning programs offered in a new network of Regional Labor Development and Training Centers.43 Another approach is to encourage the provision of lifelong learning opportunities through employer-provided or privately provided training as in the French vocational training levy. This method can mobilize significant financing, but it entails the disadvantage of raising employer costs and thus discouraging job creation. A less interventionist and more 42 Halasz, Gabor, and Herbert Altrtchter. 1999. "Comparative Analysis of Decentralization Policies and Their Results in Central European Countries." National Institute of Public Education, Budapest 43 These centers were developed under the Ministry of Labor with the intention that they would become self- managed and self-financing. This goal has not been fully achieved. With the dissolution of the Hungarian Ministry of Labor in 1998, these centers were temporarily assigned to the Ministry of Education. This move entails a risk of compromising the centers' close links with employers. -46 - Successfully Integrating with the European Union employment-friendly approach is the Anglo-Saxon model of tax incentives for employer- provided training. This model works well in the United States, where employers invest heavily in training in order to remain competitive. Financing Quality Higher Education Higher education needs to continue to develop in order to meet the needs for investment-driven growth and to support the CEECs' move toward innovation-driven growth. For this purpose, higher education programs need serious investments to improve quality. Improved laboratories and stocks of materials play an important role in quality improvement for the sciences. More generally, major investments are needed to update and restock university library facilities with up-to-date journal collections and computer-based research tools. These investments to diversify sources of information are crucial if higher education systems in the CEECs are to gain access to the latest developments in knowledge and to support a more inquiry-based form of teaching and learnuing. These investments are also costly. In a situation of constrained budgets, how can the CEECs afford to make these improvements while enrollments continue to expand? The first candidate for action is more intensive use of teaching staff through increasing minimum teaching hours so that they are more consistent with standards in the OECD countries. Another simple measure, adopted by all the CEECs, is to allow higher education institutions to generate and retain income through the provision of professional services. This can help develop links between academia and the economy, but it needs to be implemented judiciously in each institution order to prevent the income-earning role from displacing the roles of education and research. Another measure to mobilize financing is cost recovery in the form of student fees. There are widely varying approaches to the collection of student fees in higher education. The Czech Republic has resisted imposing student fees in higher education, and the Slovak Republic allows student fees only for part-time courses (highly demanded in new areas such as IT). A basically different approach is to restrict state budget financing to either the most capable institutions or the most capable students. A number of CEECs (including Estonia, Latvia, Romania, and Bulgaria) have selectively applied student cost recovery under which students who score highest on the university entrance examinations are financed by the state budget, while students who receive lower scores (but still high enough to gain entry to their preferred institutions) are admitted as "contract" students, and pay fees intended to recover the full cost of their education. This system of cost recovery for contract students is equivalent to operating lower-quality private universities within public universities. Where the number of contract students is large enough, it sometimes takes the form of separate classes for contract students. Cost recovery through student fees tends to discourage attendance by capable students from poor households. This problem can be mitigated, but not eliminated, by combining student fees with a means-tested scholarship scheme or a student loan scheme. But in order to perform this role effectively, student loan schemes need either budget-financed interest subsidies or effective linkage to future earning streams (as attempted by Hungary). Neither of these approaches addresses the bias that results from the fact that students from higher-income - 47 - Expenditure Policies Toward EU Accession families are more able to afford private tutoring and the informal fees to attend better schools, and thus to obtain higher scores on their university entrance examinations and free admission to the higher education programs of their choice. An alternative is to apply budget selectivity at the institution level. This amounts to triage: Public budget financing is restricted to the leading universities that are financed amply as centers of excellence, with other institutions of higher education finding their own financing and establishing their own standards of quality. This can be an effective approach, as the U.S. model of parallel public and private higher education institutions demonstrates, but an effective accreditation system is necessary for this model to function without creating false expectations on the part of the public. Accreditation should play a key role in monitoring quality and establishing accountability for higher education institutions. Accreditation provides the metric by which students, parents, and the institutions themselves can judge their performance. Most of the CEECs aspire to develop an accreditation capacity as an instrument of quality assurance for both public and private higher edtucation institutions. This work is at an early stage in most of the CEECs. Romania has recently developed accreditation standards for higher education institutions that led to the closure of a number of the private universities that had been created during the 1990s. Concluding Remarks The CEECs need to speed up the redeployment of expenditure in response to the demand for structural changes in education arising from the passage of the "baby bust" and the new skills requirements of the global economy. The feeble incentive for efficiency conveyed by the prevailing input-based financing formula and, in many countries, the small size of local jurisdictions in charge of schools have hindered the process. As a result, both the generally high quality and the accessibility of education are beginning to slip away. New policy patterns are emerging, however, that emphasize composite capitation formulas for school financing at the national level and the assignment of education management to larger jurisdictions. Disappointing results on standardized international aptitude tests have also called attention to the need for education systems to better develop the higher order of skills needed for most of the fastest growing jobs in the global economy. National governments will need to exercise greater leadership in bringing about curriculum reform (as in Slovenia), to consolidate programs of vocational education into fewer specializations for dusters of kindred occupations (Hungary providing a good example), and to define appropriate mechanisms to finance the quality higher education that is needed to support the CEECs' move to innovation-driven growth. This may involve concentrating state budget financing on either the most capable institutions or the most capable students, and letting (duly accredited) private institutions, financed through tuition payments, look after the rest of the sector. - 48 - Successfully Integrating with the European Union UPGRADING ENVIRONMENTAL STANDARDS In seeking to join the European Union, the CEECs are also binding themselves to much higher environmental standards than the ones they have been enforcing. While the CEECs are looking toward better amenities and a better quality of life, the magnitude of the investments involved has given them pause. Are such investments economically justified? Can the investment and also the maintenance costs be financed? Can such outlays be made consistent with the macroeconomic programs-laid out in the PEPs? Benefits of Improved Environmental Standards Meeting the enviromnental acquis will bring considerable improvements to people's lives. The European Commission (EC) has recently quantified the expected benefits for the CEECs.44 The results of the exercise are given in table 3.9.45 The expected benefits from meeting air standards are generally highest followed by those arising from stricter water standards, especially if the upper estimates are taken (the exceptions are Bulgaria and the Czech Republic). Table 3.9 Total Benefits over the Period 2005-20 (billion euro) Water Air Waste Total Benefits Per capita Low High Low High Low High Low High euro Bulgaria 1.58 4.20 1.07 11.00 0.20 6.62 2.85 21.82 3,010 Czech Republic 15.23 24.05 7.10 35.10 0.93 11.20 23.26 70.35 9,090 Hungary 2.72 10.49 5.74 39.92 1.12 18.50 9.58 68.91 7,770 Poland 13.59 31.96 25.80 149.90 1.60 26.30 40.99 208.16 6,440 Romaria 3.96 12.15 7.59 56.95 0.83 26.30 12.38 95.40 4,790 Slovak Republic 3.00 6.61 3.40 21.90 0.29 4.28 6.69 32.79 7,310 Slovenia 1.47 3.44 0.68 4.62 0.24 2.82 2.39 10.88 6,640 Baltics Estonia 0.26 0.99 0.39 2.05 0.09 1.75 0.74 4.79 3,950 Latvia 0.38 1.34 0.49 3.12 0.05 1.07 0.92 5.53 2,690 Lithuania 1.23 2.75 1.56 7.98 0.06 2.00 2.85 12.73 4,210 Total 43.42 97.98 53.82 332.54 5.41 100.84 102.65 531.36 6,060 As percent of total 42.00 18.00 52.00 63.00 5.00 19.00 100.00 100.00 Note: Net present value at a 4 percent discount rate. Source: Ecotech. 2001. The Benefits of Compliance vith the Environmental Acquisjfr the CEECs. Brussels: European Commission. 44 The study covers all 13 candidate countries: Bulgaria, Cyprus, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Romania, Slovenia, the Slovak Republic, and Turkey. 45 Benefits covered in this exercise include health, resource, and ecosystem protection and the directives covered include al those with an Impact on air, water, and solid waste. The results from the exercise are reported as the net present value of benefits, discounted at a real rate of 4 percent. In actual practice, countries may use higher discount rates, if only to ensure that they prioritize those investments that wiU yields earlier benefits for the populations. - 49 - Expenditure Policies Toward EU Accession However, estimates range widely across categories and countries.46 The greatest uncertainties are associated with the potential benefits of EU waste treatment standards, followed by those of air and water standards. Some of the variations across countries, on the other hand, result from differences in baseline environmental conditions. For example substantial benefits will result in Bulgaria and Estonia from improvements in drinking water connections. A large number of households are not connected to piped water. The Czech Republic has the greatest river length of all the CEECs (76,000 km). At the same time, not a single river is of "good" quality (by EU standards); that is, 10 percent are of "fair" quality, 10 percent of "very bad" quality, and the remaining 80 percent are of either "poor"' (40 percent) or "bad" (40 percent) quality. Compliance with EU water directives will improve this situation considerably: 10 percent are expected to be of "good" quality, and all rivers of "poor," "bad," or "very bad" quality are expected to improve to fair quality after successful implementation. This yields substantial benefits to the Czech Republic, giving much higher figures than those for some of the other countries. Cost of Complying These benefits will unfortunately come at a cost. Indeed, discussions on the environmental acquis often revolve-around the cost of meeting the environmental requirements for accession. The EC took the first stab at this in 1997, and these figures have since been updated and revised. Table 3.10 provides its initial estimates of the investment costs. By any standard they are large; at f 110-116 billion over a period of some 20 years, they amount to between 23 percent and 134 percent of the countries' present GDP. In per capita terms, the total investment amounts to between e 760 and £ 1,760. Table 3.10 Estimated Costs of Environmental Acquis for the CEEC Water Air Waste Total Investment Per As Water Waste capita Percent of supply water Total Total Min. Max. Min. Max. (euros) 1999 GDP Bulgaria 2.20 2.70 4.90 5.10 1.80 5.10 11.80 15.10 1,668 134 Czech Republic 2.20 1.10 3.30 6.40 8.00 3.80 17.70 13.50 1,427 32 Hungary 3.50 3.10 6.60 2.70 2.10 4.40 11.40 13.70 1,306 32 Poland 4.40 13.70 18.10 13.90 2.20 3.30 34.20 35.30 927 26 Romania 3.80 6.30 10.10 9.10 1.00 2.70 20.20 21.90 943 72 Slovak Republic 1.00 0.90 1.90 1.90 0.30 1.60 4.10 5.40 760 23 Slovenia n.a. n.a. n.a. 0.69 1.15 1.15 1.84 1.84 n.a. n.a. Baltzcs Estonia 0.13 1.38 1.51 Latvia 0.11 1.60 1.71 8.45 0.45 0.85 8.90 9.30 1,148 48 Lithuania 0.11 2.27 2.38 Total 17.45 33.05 50.5 48.24 17.00 22.9 110.14 116.04 As percent of Total 14.00 27.00 42.00 40.00 19.00 100.00 Notes: Per capita and GDP percentage numbers are for averages of minimum and maximum investment (billion euros unless otherwise indicated. n.a.: not avallable. Sources: EDC (1997); IFO Institute (1997). 46 The high estimates vary by as much as an order of magnitude as the low estimnates, reflecting the substantial uncertainties surrounding environmental benefit estimation in general. - 50 - Successftdly Integrating with the European Union Estimates have also been made of the extra annual operating costs that will be incurred. These are of the order of euro 8-12 billion, or about e 80-120 per capita for the group as a whole. Dividing the expenditures equally over 20 years gives an annual spending of between 1.2 percent and 6.7 percent of present GDP depending on the country. This calculation suggests, therefore, that on average about 5 percent of GDP will have to be devoted to environmental related expenditures, which is more than double the levels of expenditure on these items in CEEC countries in recent years. Comparison With Cohesion Group Countries Comparable estimates were made at the time the cohesion group (Spain, Portugal, Ireland, and Greece) joined the EU. Compared to the latter, the CEEC estimates mentioned above are about 50 percent higher on a per capita basis, with most of the difference coming in the costs of treatment of wastewater and drinking water (EDC, 1997, "Compliance Costing for Approximation of EU Legislation in the CEEC," European Commission: DG Environment, Brussels). This is possibly due to a much poorer starting position for the CEECs with respect to the management of these services. It is important to note, however, that the cohesion group countries did not fully comply with the directives in accordance with the agreements they made at the time of entry, and indeed there are still some areas of noncompliance in these countries. Some useful lessons can be learned from the experience of these cohesion countries for environmental management in the CEEC countries. One key lesson is that fulfilling the acquis may in practice take considerably longer than the timetable to which the countries are agreeing. The other lesson is that the nature of the environmental problems may well change dramatically after accession, with transport and packaging the areas of most rapid growth. In Greece, for example, the length of motorways increased more than 14 times between 1980 and 1987 and the number of motor vehicles increased almost three times. This has consequences for air pollution, especially for particulate and volatile organic compounds. Also observed are large increases in the amounts of industrial and household solid waste, especially hazardous waste (Kuik and Oosterhuis, 2001, "Lessons from Southern Enlargement of the EU for the Environmental Dimensions of Eastern Enlargement, in particular for Poland", Nota di Lavoro, 59.2001, FEEM, Milan). Relating Benefits to Costs Be that as it may, it is neither the level of expected benefits nor the level of expected costs that should primarily receive our attention, but rather the comparison of the benefits against the costs. The good news is that the ratios involved (table 3.11)47 are generally greater than 1, but not always. In particular, for the waste directives the estimates are below 1 in all cases including the "low" benefits and even in some cases with the upper end estimates of benefits. In the case of water, the ratios are less than 1 for some countries in terms of the low 47The table has been constructed by annualizing the benefits in table 3.9 at a real discount rate of 4 percent over 15 years. This figure has been divided by the sum of the annualhzed 20-year investment estimates in table 3.10, and the annual operating costs. - 51 - Expenditure Policies Toward EU Accession benefits (Bulgaria, Hungary, Romania, Estonia, and Latvia). In the case of the air directives there is only one case of benefits below 1 (Bulgaria). These estimates must, however, be interpreted with some care. First, an overall ratio of greater than one does not imply that (a) the net benefits of all projects will be greater than one, (b) the ratio cannot be increased by delaying the implementation of some of the directives in some places, and (c) the distributional and social costs of implementing the directives may outweigh the assessment based on the ratios presented. Indeed, for the reasons given in this note, that is likely to be so in a number of cases. It does point, therefore, to the need to carry out a careful benefit/cost analysis of the implementationfur each investment. Table 3.11 Ratio of Benefits to Costs of Complying with Environmental Directives Water Air Waste Total Benefits Low High Low High Low High Low High Bulgarna 0.73 1.94 0.26 2.64 0.08 0.40 0.20 0.85 Czech Republic 10.43 16.48 1.36 6.70 0.08 0.91 1.06 2.92 Hungary 0.93 3.59 2.60 18.07 0.36 1.30 0.63 9.16 Poland 1.70 3.99 2.27 13.18 0.49 2.46 0.98 13.12 Romania 0.89 2.72 1.02 7.65 0.56 3.01 0.51 9.11 Slovak Republic 3.57 7.87 2.19 14.09 0.66 0.83 1.37 10.55 Slovenia n.a. n.a. 1.20 8.18 0.14 0.76 1.06 4.45 Balhcs Estonia 0.39 1.48 Latvia 0.50 1.77 8.08 49.71 0.30 3.85 0.28 1.27 Lithuania 1.17 2.61 Total 1.94 4.39 2.41 14.89 2.67 13.51 0.73 3.23 Source: Own calculations based on previous tables. How Trade-Offs Can Be Improved The total cost figure is valuable only to give some idea of the overall size of the task. What is more important for the countries is to prepare detailed plans, covering periods of three to five years, that ensure compliance with the agreements arrived at under the environmental chapter and to do so in a way that minimnizes the costs as well as ensuring that the underlying financing arrangements are sustainable. Indeed, while the acquis communautaire is prescriptive on environmental standards, it leaves considerable latitude on how to meet them. The price of complying varies accordingly. By way of illustration, table 3.12 gives some more recent alternative estimates, partly based on the World Bank's work and partly on the firming up of estimates by other parties that are considerably lower than the earlier ones. The more detailed studies such as the ISPA strategy for Lithuania come up with figures that are as much as 75 percent lower than the earlier calculations. The World Bank's Bulgaria study gives figures about half those of the earlier EC study. - 52 - Successftlly Integrating with the European Union Table 3.12 Alternative Estimates of Investment Costs for Compliance (billion euros) Country EC 1997 estimate Other estimates Bulgariaa 11.8-15.1 5.5-8.0 Polandb 34.2-35.3 17.0; 19.0-39.0 Lithuamac 2.38 0.64 a. Based on Bucknall, Cestti and Damianova, 2001, "Bulgaria: The Challenges of Complying with EU Enviromnental Directives," ECSSD Working Paper 30, World Bank. b. Euro 17 billion is from a second EC study, TME (Institute for Applied Environmental Economics), Cost of Compliance with EU Environmental Directives in Poland: Paper for Seminar 10-3-1999 at the Polhsh Ministry of Environmental Protection, March 1999. Range of euro is from World Bank study (Gordon, Hughes, and Julia Bucknall. 1999. "Poland: Complying with EU Legislation." ECSSD Working Paper 14, World Bank, Washington, D.C.). c. Estimate is only for water. "Other estimate" is based on Lithuania National ISPA strategy (2001). Savings can be achieved in a number of ways:48 * Following a least-cost investment plan, especially in energy-related investments. This in turn can be promoted by the use of economic instruments such as bubbles and permit trading. * Efficiency with which municipalities make investment decisions. Opening up procurement to international tender and undertaking careful project appraisal in evaluating the design of the schemes will reduce costs significantly. * Designing the investments to take account of the lower demand for some services when future service charges will have to recover capital and operating costs. The World Bank has had stark experience of this in the wastewater projects it funded in the Baltic states, where the level of capacity is turning out to be substantially in excess of demand as a result of the large increase in volume based charges (box 3.1). 48 Total costs may also fall over time, as manufacturers drop their prices for capital equipment in response to larger level of production (Hager, 2000, 'Environmental Investment in the CEEC preparing for Accession", mimeo, World Bank). - 53 - Expenditure Policies Toward EU Accession Box 3.1 Bank Experience with Water Utilities in the Baltic States The World Bank has provided financing for five water and wastewater projects in the Baltics: one in Estonia, two in Latvia, and two in Lithuania, with a total cost of US$134 million. The upgraded plants are now in operation and the wastewater meets Helsinki Commission standards (HELCOM protects the Baltic marine environment as mandated under the Helsinki Convention) and the drinking water meets the highest standards. While the projects have been successful with respect to many criteria, one of the most serious problems has been the drop in demand for water and the generation of wastewater. The projects were planned assuming that the demand would stabilize at a moderate level of consumption similar to that experienced in other comparable countries. In fact this has not materialized and the actual level has been as much as 50 percent lower than anticipated. This extreme decline in demand has led to: * Over dimensioning of systems • Over investment in systems o Higher than optional O&M cost • Water quality problems in potable water networks The key lesson learned is that, while drafting the financial and economic projections, the project teams should be conservative in their assumptions of critical variables, which have the highest impact on revenues and costs. These tend to be too optimistic, and in the long run the projects are likely to experience lower than anticipated economic and financial rates of return. Project appraisal should ensure that the assumptions are realistic and that contingency plans are drawn for coping with deviations from the most likely outcomes. The range of possible outcomes with their respective probabilities, although subjective, should be listed in the appraisal document. Price escalations also tend to be greater than expected (in real terms) owing to elimination of existing market distortions and price disequilibria, which were present in the economies during the early transition process. ThLs is also an important reason for the dramatic reduction in water sales. Tariffs have had to increase more than anticipated. * New investments are likely to reduce total operating costs, because the operation and maintenance costs of new equipment will be lower than that of the equipment it replaces. This gain has not been fully accounted for in the above figures. * Present value of total costs will fall if the more expensive items are scheduled further away in time. This approach would be all the more justifiable when the ensuing benefits are comparatively low. * Environmental mitigation costs may not expand at the same pace as income. With growth and convergence, countries should therefore be better able to meet the acquis. The demand for services such as energy and transport, and the related costs, are likely to rise in tandem, but the demand for services such as water may not increase proportionately. The resulting increase in total costs may therefore well be less than the growth in national income. * Relaxing statutory standards (particularly for water treatment), if it can be shown that the investment is seriously uneconomic (for example, if the community served will decline dramatically in the next few years) and that the savings in costs by reducing the attainment level by a small percentage will allow more plants to be - 54 - Successfully Integrating with the European Unwon built and a higher overall contribution made to meeting the ambient environmental goals.49 Negotiations on the Envirornment Chapter Those countries that have closed the environment chapter have made commitments to complying with the acquis by agreed dates. The highest priority in all cases has been given to legal approximation, ensuring that the national legal framework is consistent with the EU legislation. Next in priority has been the institutional strengthening of supervisory bodies and environmental agencies. All this is expected to be completed before accession. The investment program necessary for compliance has a completion date of around 2015, with interim targets for key directives. For example, in the case of Lithuania the directive on the sulfur content of petrol and diesel has to be complied with by 2005, the Large Urban Wastewater Directive by 2010, and so on. At the start of the negotiations the country asked for eight transitional periods, but in the course of negotiations it withdrew its request for five of them and shortened the other three. Even with these transitory periods, the agreed schedules may be difficult to implement. This is principally because a large part of the accession negotiation took place before, or concurrent with, the implementation of the benefit assessment study mentioned above. There was also probably a feeling that the environment chapter could be renegotiated later. In that event, a more careful assessment of the costs and benefits as outlined here could play a useful role. Funding Environrental Investments Although a number of CEECs have now concluded the environmental chapter of the negotiations, it is fair to say that they have not fully established how these investments will be funded. Indeed, there are ongoing concerns in the front line CEEC countries about the availability of co-financing from the national budget to match the EU accession funds available under ISPA, PHARE (Assistance for Economic Reconstruction), and SAPARD (Special Accession Program for Agriculture and Rural Development). Indeed, the slow rate of development of projects for such funding can be partly attributed to this factor (as well as to the lack of administrative capacity to implement the acquis). The problem that many countries are having in meeting the local cost share of investments is unlikely to be solved merely by increasing the allocation from the public budget for the environment. Other solutions must also be found. In most of the CEECs one source of public sector finance for the environment has been through a number of environmental funds. These funds receive their payments from pollution fees and charges, including payments for noncompliance. The funds are used to pay for environmental services (such as monitoring and research and development) as well as for some investment in pollution abatement at the enterprise and municipal level. While they have 49 Marginal costs of standards rise with the standards themselves, and it holds par excellence in the areas of wastewater treatment. - 55 - Expenditure Policies Toward EU Accession played an important role in some countries in the post-transition period, there have been some doubts as to the efficiency with which funds are allocated, and the future does not look promising for them as a vehide for environmental finance. Principally, this is because their consistency with the EU institutions is not clear. Poland, which has closed the environmental chapter in its negotiations, is looking into environmental funds carefully. The concern is that enterprises will seek legal exemption from payment of "pollution charges" if the same changes are not imposed on their competitors in the EU. If this challenge is successful, then the revenue source of the fund will disappear. Since earmarking of this kind is not favored as a means of financing public expenditures (with the exception of payments for specific services, such as monitoring) and is rarely used in the EU, further difficulties can be expected with its continued existence in the CEECs.50 The scope for moving some items out of the public budget is also quite large.51 As far as investment in the manufacturing industry is concerned, the more of the sector that is privatized, the less will be the burden of the acquis on the public budget. In this regard, experience has shown that successful privatization of some of the larger and more polluting industries requires a clear understanding of the liability for past environmental damages, through an internationally acceptable audit, accompanied by a legal agreement with the government as to the new owner's responsibility for clean-up. If internationally credible investors are to be attracted into bidding for such enterprises then these issues have to be addressed. Furthermore, the state's responsibilities have to be backed by a credible program of investment in remediation. Otherwise the uncertainty on the private party's side will result in a failure-to bid, or in an offer that reflects the increased risk. This illustrates a situation in which it may pay for the state to undertake investment in remediation with a view to making the privatization effort more successful and generating higher revenues from the sale of state assets. The same is true to a large extent of public utilities and infrastructure. A Czech study showed that as much as half the big-ticket items could be shifted out of the public budget under certain assumptions.52 To date, however, the detailed national plans for the adoption of the acquis paid limited attention to the role of the private sector, and there has been mixed progress on the ground. Hungary has totally transferred power generation to the private sector, while Poland moved 9 of its 27 generation stations into private hands, and other countries moved none.53 Fortunately, as noted in chapter 1, the movement is gathering pace. 50 From the perspective of the international financial institutions, extrabudgetary earmarked funds are not encouraged. The IMF and the World Bank have almost always asked for the eco-fund to be part of the consolidated budget. 51 The increased level of private sector activity implies a greater effort by the state to ensure compliance. This in turn requires mvestment in capital equipment for monitoring and testing, etc. Funding for this can be obtained from ISPA, if the demands can be bundled to meet the 65 million threshold. 52 Czech Republic, "Towards the Accession", Main Report, A World Bank Country Study, The World Bank, Washington, DC. 1999. 53 A comprehensive status of privatization in the energy and utility sectors for the CEEC countries does not appear to be available and is being prepared by the Bank. - 56- Successfully Integrating with the European Union Privatization is not, however, the only way of sharing the burden of upgrading environmental standards. Another is to commercialize the enterprises even when they are nominally state-owned, and to raise the finance for the investment through commercial loans. This strategy is being followed widely in all the CEEC countries. It was successfully adopted, for example, by Poland in the power generation sector. Even before the privatization of the nine units referred to above, the financing for investing in pollution control equipment came from commercial loans, backed by power purchasing agreements between the commercialized generating units and the state electricity authority. The same strategy has been adopted in municipalities dealing with water supply and wastewater in a number of countries, including the Baltic states. Although it is successful in taking the direct investment cost off the budget, this approach suffers from the problem that the borrowing is, invariably, guaranteed by the government and therefore forms part of the consolidated national debt. There is also the issue in such cases, however, of subsidies to these enterprises through working capital and loans from state-owned banks, given at below commercial rates. To the extent that these practices prevail (and they are still quite common), the institutional mechanism of commercialization for taking environmental costs off budget will still leave some budgetary burden. The commercialization of utilities has, however, opened the door to raising charges for their services to a level at which at least the costs of the new investments are recovered. This will result in reduced levels of demand, thus saving the budget the cost of the subsidies they receive on ongoing operations. It will also reduce the size of the investment needed (box 3.1).54 Another way in which commercialization can save costs is through a rationalization of provision, as many utilities are currently too small to make cost-effective investments or manage operations in an efficient manner. Concluding Remarks Complying with the environmental acquis will generate high benefits (the expected benefits of meeting air standards are highest, followed by those associated with water standards). The costs involved, however, can be high. While benefits are generally expected to exceed costs, this is not always the case. Each individual investment will therefore need to be subjected to a thorough cost/benefit analysis. A number of measures can be taken to improve the related trade-offs. These nieasures include adopting least-cost investment solutions (especially for energy-related investments), opening up procurement to international tender, and phasing in the more expensive compliance measures over time. Funding the local cost of environmental investments is not merely a matter of increasing expenditure commitments. Rather, there is a need for administrative changes in budget implementation to make funding more efficient (for example, consolidation of small ecological funds). Commercialization could potentially be helpful, but it still entails some budgetary burden. A more effective means of furthering EU compliance is likely to be through the s4 The constraints on raising charges, however, are real and raise issues of affordability for the poorer customers. Most countries adopt some lifeline rates to get round this, although this tends to be done in an ad hoc fashion and is not based on a careful assessment of a structure that meets specified targets at least cost. - 57 - Expenditure Policies Toward EU Accession privatization of some of the larger and more polluting industries in CEEC countries. A clear understanding of the environmental liabilities of past damages is necessary and a legal agreement with the government explaining the new owner's responsibilities is crucial to attract credible international investors. UPGRADING TRANSPORT NETWORKS To spur growth, the CEECs are also called upon to upgrade and redirect their transport networks to the requirements of European integration. We will discuss roads and motorways because they take the lion's share of the investment programs.55 Key challenges include the following: * Rapid increase in motorization and drastic shift in modal split Over the past decade the demand for road transport has grown radically, owing to the sharp decline of rail transport and unprecedented growth of passenger car ownership. The number of vehicles in the CEECs increased on average between 70 and 120 percent in recent years. The supply side, however, has not kept pace. The road sector has been a victim of under investment in maintenance and modernization. As a result, the roads are mostly congested, slow, polluted, and unsafe. * European integration challenges: The most important EU accession challenges arise from the liberalization of international trucking, the increase in axle loads to the higher EU standards (11.5 t/axle), and the accelerated corridor development programs under the Trans-European Network program. Furthermore, approximation with the EC transport legislation and the set up and strengthening of the EU-compatibre institutions necessitate further changes of the public administration in the road sector. * Expanding too rapidly: Although in most countries the roads needed to stimulate domestic economic growth are largely the same as those needed for international transit traffic, the attempts to expand the motorway network too rapidly and to standards that are not always economically justified have already diverted funds from maintenance and development. Indeed, even in the main corridors, the rapidly increasing volume of heavy truck transport has not been matched by a corresponding expansion of capacity or strengthening of the pavement. Road Networks as National Assets Around the world road networks are increasingly being seen as national assets that need to be managed as any other asset portfolio.6 It is in that spirit, for example, that the Japanese public highway agency manages assets about the same size as General Motors, while the U.K. 55 The TINA study identified transport investnent projects on the Pan-European corridors, the cost of which would reach 2 percent of GDP. Other investments in local urban and rural areas, as well as those serving regional mtegration within the country, would come in addition. 56 Following the suggestion of the U.S. Federal Highway Department and its Office for Asset Management the colloquial terms of asset and equity are written with capital letters, if considered as technical terms regarding the economic accounts. - 58 - Successfully Integrating with the European Union highway agency (a relatively smaller road agency responsible for just over 10,000 km of roads) manages assets representing around the same value as IBM and AT&T. In the CEECs, the replacement value of the 493,682 km of main roads may amount to as much 32 percent of total GDP in the 10 countries combined, ranging as high as 90 percent of GDP in Estonia. Table 3.13 gives details of main roads (motorways, first class and secondary roads, excluding municipal and private roads) and their replacement value57 by country. Table 3.13 Replacement Value of Main Road Network, 1999 Main road network Replacement value Replacement value as Country (km) (mill. euro) (percent of GDP) Bulgaria 37,286 9,106 74 Czech Republic 55,408 14,039 26 Estonia 16,434 4,644 89 Hungary 53,267 13,599 27 Latvia 20,328 4,550 72 Lithuania 21,312 5,614 54 Poland 173,048 40,746 24 Romania 78,615 20,396 58 Slovak Republic 17,807 5,209 24 Slovenia 20,177 5,351 25 Total 493,682 123,252 32 Note: Excluding municipal and local roads. Source: World Bank. These assets are not being properly maintained. As is illustrated in figure 3.1, owing to insufficient funding, road asset conditions are generally fair to poor. This is either because of continued deferred maintenance or low productivity, or both. The rural and urban roads are usually the ones that suffer from under-maintenance. In some countries, however, maintenance is below standard also on the national road network simply because of lack of funds. Figure 3.1 Condition of Road Network, 1999 Bulgana Czech Republic _ i_ Estonia Hungary _. _rr _ Latvia , . Lithuania ir,.: . - -Fir Poland . -*- . *Poor Romania _ Slovak Republic Slovenia _ AVERAGE - , 0% 20% 40% 60% 80% 100% Source: NEI transport consultants. 57 Estimated at E2.5 million /km for motorways, e0.5 milion /km for national roads, E0.2 mnillion /km for secondary roads. -59 - Expenditure Policies Toward EU Accession In addition to the poor conditions of the existing roads, their throughput capacity cannot cope with the sudden increase in traffic. Selective network upgrading as well as new constructions over time are warranted. Mostly driven by a desire to join the EU and generate business for domestic contractors, all CEECs initiated an ambitious program to develop their road and particularly motorway networks too quickly at comparatively high costs. Few motorway sections have sufficient level of traffic to justify the investments. As seen in figure 3.2, few countries have average daily traffic densities (AATD) approaching the average 12,000-15,000 benchmark typically required to justify motorway construction (on flat land). By comparison, the AATD on Spanish motorways is about 30,000. Fige 3.2 Average Daily Traffic Density on Motorwa s and Two-Lane Major Roads 25000.,c = 20000., 15000 ' 5000 i Bulgaria Czech Hungary Lithuania Poland Romania Slovak Ukraine Spain Republic Republic | OMotorway 8 2-lane major road Source: OECD, 2000, " Road Strengthening in Central and Eastern European Countries," Report, Paris. As a result, the attempt to expand the existing network in a relatively short period of time (see below) is displacing higher priority maintenance expenditure. As a result, actual expenditure on average falls short by 1.5-2 times below normative maintenance costs. In a portfolio approach, road expenditure should be prioritizes according to economic principle. A general guideline is to compare the rate of return of the different road works both within the same category of works and among the different investments. Table 3.14 presents some benchmark numbers in this respect.58 Table 3.14 Internal Rate of Return on Road Projects: Best Practices Average Type of road project Denomination internal rate of return Road niaintenance projects "Blackfield projects" 40 percent and more (depending on state of road deterioration) Road rehabilitation projects on "Brownfield projects" 20 percent existing roads New road construction projects 'Greenfield projects" 10 percent (World Bank requirement is often 12 percent or above) Source: World Bank. 58 "Contribution to the IRF: Seminar for the-Transition Countries," Bucharest, 2001. - 60 - Successfully Integrating with the European Union This implies that from an economic point of view, the following rules of thumb generally apply: (a) execute maintenance projects first, (b) undertake the rehabilitation and upgrading projects (in the CEECs they are needed to convert the road network to the bearing capacity for the heavier HVGs with higher gross weight and axle load factors), and (c) consider new construction projects. Where new construction is required to build missing links and to align the road network with the requirements of European and subregional integration and of the restructured domestic production, we could argue for staged construction, because such roads often need not be up to the highest road standards (for example, a mnissing road link may better be established as a two lane road at first and not as a four lane motorway or expressway). The Slovak Republic has embraced such an approach. The country had originally embarked on an over-ambitious motorway program shortly after 1993. The length of the proposed motorway network took no account of the cost feasibility or of the scale of the benefits that might be achieved. Since the budget did not have the required resources, the government took many long-term, but also short-term, loans, which increased its contingent liabilities. The present administration has substantially revised the motorway program, bringing it more in line with what is financially feasible, mostly by reducing the design standards, slowing down the rate of construction, building in short sections that have acceptable economic rates of return, and building most new sections as a single carriageway (stage construction). Financing Road Investments The CEECs have followed one, or both, of the following financing models: involvement of private sector financing through concessioning and/or traditional public budget financing. Private Sector Concessions While the first option holds much promise, the early experience with concessioning has unfortunately not been encouraging. Hungary and Poland were among the first countries to fund motorway construction through public-private partnerships (concessioning). In Hungary, the private motorway concessionaire that won the first "Build-Operate-Transfer" contract in the region went nearly bankrupt, forcing the Hungarian government to "purchase" back the motorway (Box 3.2). The experience in Poland has hardly been more encouraging. - 61 - Expenditure Policies Toward EU Accession Box 3.2 Hungary M1/M15 Motorway Concession Project The M1/M15 project was the first toll motorway project tendered and implemented in Central Europe. Following a successful tender and financing, construction was largely completed on time and within budget even though the construction period was ambitious and Hungary underwent a period of high inflation. Ml traffic at opening and traffic growth during the first three years was substantially below expectations, resulting in the impossibility of servicing debt. The level of toll rates turned out to be politically unacceptable, and a court case made the financial situation untenable. Attempts to restructure company finances, starting with the issue of letters of credit by the government and shareholders, remained unfruitful. The government and lenders agreed on a substitution process after three years of operation. The government brought this vital piece of infrastructure back into Hungarian state ownership while accepting part of the debt Tolls were reduced (and abolished) and a vignette system prepared. Taking on the Ml debt also meant that the motorway construction budget for at least one year was completely exhausted. In addition, private funding sources for the road sector seem to have become more cautious.. Of the ambitious motorway program of Hungary outlined in 1991, only parts were realized by the end of the century and the early completion of other parts remains doubtful. The M1/M15 experience shows that sound traffic projections and prudent economic evaluation of projects are essentiaL While the lack of adequate traffic was the key reason for the private concessionaire's failure, the complexity of the issues and the confidence lost made the Hungarian government decide against trying to find a viable PPP solution. With the benefit of hindsight a positive lesson from the first BOT (Build-Operate-Transfer) project in the region is: * Construction was completed on time and within budget * Operation and maintenance during the short period thereafter were effective and on highest standard. * During the critical economic period following its opening to the west, Hungary benefited from the Ml while not contributing to its financing. Many potential factors that can cause private funding to fail include very weak institutions, almost nonexistent regulatory capacity and local financial market capacity, a judicial system in the process of transformation to a market economy, and generally unstable economic situations and as result high perceived risks. In addition, the equity share of the operator is usually very low, which will not ensure that the concessionaire/operator would be sufficiently committed to the contract dauses. Paying tolls is a new phenomenon for the people in the CEECs. As a result, when the level of the toll is too high, public acceptance is critically low. In the experience of Hungary and Poland, the main cause of initial failures was the lack of realism of the underlying traffic forecasts. In the case of the "re-nationalization" of the M-1 motorway in Hungary the forecasted increase in traffic never materialized. In some cases, however, the government is willing to cover the difference between the toll level acceptable to the public and the one required by the concessionaire to cover for operating costs. Undoubtedly, this system of "shadow tolling" can raise further doubts about the financial sustainability of the approach. In both Poland and Hungary, the revised motorway expansion programs deserve credit for showing the ability to learn from the recent lessons (box 3.2). - 62 - Successfully Integrating with the European Union The "Other' Option Public (Budget) Financing Most of the CEECs continue to finance their motorway expansion programs from public resources. They typically spend 0.5-1.5 percent of GDP on roads; this compares to 1-2 percent of GDP in the EU.59 As noted above, a high proportion of the road expenditure (around 0.9 percent of GDP) goes to the national roads (motorways, expressways, and the rest of the national road network), leaving regional, rural, and urban road networks behind and less than half (on average 40 percent) of the total road budget is used for maintenance. Directly or indirectly, road users fund these expenditures. Road User Charges While the structure of road user charges in the CEECs generally corresponds to the practices of the EU countries (table 3.15 and box 3.3), the level of taxes, particularly the fuel and vehicle taxes is nevertheless lower. Fuel taxes are around two-thirds of the EU on average. Fuel taxes in Slovenia, Hungary, and the Czech Republic60 are above the EU minimum,61 whereas the lowest fuel taxes are in Romania, Lithuania, and Bulgaria. With regard to vehicle registration taxes, the Czech Republic, Slovenia, and the Slovak Republic represent higher taxes than the EU minimum, while Romania, Lithuania, and Bulgaria are lower. Table 3.15 Tolls and Vignettes in the CEECs Toll roads Vignettes and transit charges Bulgaria No toll roads, Transit charge for some non-EU foreign trucks and buses have to pay at except for bridges borders for use of motorways and three-quarter lane roads: EO.1-0.4/km to Romania Czech Republic - Vtgnette for use of motorways and expressways annual fee is time related (10 days, month, year) since 1999. Estonia - But the Baltic vignette has been under consideration Hungary Toll roads: 3 of 4 Vignette system can be used as alternative for paying tolls for a period motorways are (day, week, month, year). tolled Latvia - But the Baltic vignette has been under consideration. Lithuania - But the Baltic vignette has been under consideration. Poland - Vignette for trucks and buses carrying out international transport minimum of E3.35 /day up to a maximum of E1255 /year Romania No toll roads, Transit charge for some non-EU foreign trucks and buses: eo,026 /gross except for bridges ton-km, also other taxes apply. to Bulgaria Slovak Republic - Vignette for use of motorways and expressways; foreign commercial vehicles also have to pay a time-related road tax. Slovenia Toll roads Transit charges Source: NEI transport consultants, 1999. 59 Phare study on road user charges, 1999. 60 Based on 1998 prices and taxes. 61 Which is 0,245 E/ liter of diesel. - 63 - Expenditure Policies Toward EU Accession Box 3.3 Road User Charges in the CEECs All CEECs have a system that corresponds to the EU principles [Council and Parliament Directive 99/62/EC on the charging of heavy goods vehicles for the use of certain infrastructures; Council Directive 92/81/EEC of 19 October 1992 on the harmonization of the structures of excise duties on mineral oils; Council Directive 92/82 of 19 October 1992 on the approximation of the rates of excise duties on mineral oils; Council Directive 88/77/EEC of December 1987 on the approximation of laws of Member Sates relating to the measures to be taken against emission of gaseous and particulate pollutants for diesel engines for use in vehicles [Euro I and Euro II vehicles]; White Paper on Fair Payment for Infrastructure Use of 1998). Fuel taxes typically give around 60-80 percent of total road revenues. Vehicle taxes (first registration fee and annual ownership tax) have the advantage over fuel taxes as they have the ability to differentiate among the vehicles (environmental friendliness and weight). On the other hand, their major handicap is that they are not at all related to the use of the road, or to the traffic volume. Overall, however, vehicle tax revenues are small (10-30 percent of total revenues). Several EU members and CEECs have moved away from the inherited transit fee system and introduced vignettes. In some cases they are limited to trucks; in some other countries all motor vehicles are subject to the vtgnette. The Euro-vngnette, for example (Germany, the Benelux countries, and Denmark), is required only on the motorways. This solution is hardly acceptable to the CEECs where most of the international corridors are stll not up to the motorway standards. A major concern with this type of road user charge is its loose relationship to the actual usage of the road. When they are at least time bound we can assume that once the vignettes are bought, they will be used. In addition, the revenues collected through the vignette system are rather small. Still, vignettes can serve the purpose of the transition toward more accurate user fees. Very few CEECs have introduced tolls on a network scale and m cases of private concessions the experience has not been encouraging (see earlier). Tolls levied on bridges or tunnels are much more common and meet with less resistance on the part of the users. While tolls have the closest relationshup to the actual use of infrastructure, they are often outside of the public road expenditure since as a rule they should be calculated and used to cover the maintenance and usually also the construction costs of that specific infrastructure section. Should they be used on a network scale, cross-subsidization could become possible among the different corridors. A new approach now being promoted by the EU White Paper is to bring road user charges closer to the actual use of infrastructure and at the same time to apply an integrated transport principle in which the environmental, social, safety (on average road accidents cost about 2 percent of the GDP of any, developed or developing country), etc., impacts of transport is reflected. In this way a level playing field among the modes can also be achieved. Therefore, the marginal social cost principle is recommended as the bases of road user charges The harmonized charging system for Heavy Goods Vehicles (HGVs) based on marginal social costs is planned to be mandatory from 2004, the date of the first likely EU accession treaties. The vignette system may prove to be an intermediate step toward distance and time based network tolling according to the EU recommendations. However, considerable research on the methodologies and several piloting projects are needed before this approach becomes fully operational. As the accession date approaches, these tax gaps will need to be filled,62 generating additional resources to be used potentially for the sector. A fuel price increase would hit private car users as particularly hard, because the price/wage ratio in the CEECs is on average twice that of the EU countries. Because governments will have to take action and increase taxes in the near future, the growth in private car usage may slow down. The impact of fuel and vehicle tax increases on commercial road transport operators may not be as severe as on the general public, because of the operator's current comparative advantage in their lower labor costs. However, this advantage is temporary, as the harmonization with EU working 62Unless a derogation is agreed during negotiations, as in the case of the vehicle taxes in Hungary. - 64 - Successfully Integrating with the European Union conditions (obligatory rest and drive periods) will jump labor costs. In this way any other cost increases will contribute to the erosion of the international competitiveness of the road transport operators in the CEECs. This explains the demand emanating from the road transport industry that the resources generated by the transport related taxes and charges largely fail to be channeled back to visible infrastructure improvement. Road Funds Versus General Funding There has been a long debate as to whether such expenditure should be funded out of general public resources or through the earmarking of revenues raised from road. The variety of solutions for earmarking or leaving the decision to the annual budget allocations (table 3.16) reflects the underlying dilemmna, as well as the different bargaining powers and confidence in fair treatment among various parts of the governments). Table 3.16 Earmarking of Road User Revenues ("Road Fund") in the CEECs Country State of earmarking road user revenues Road fund Bulgaria * Fuel tax: partly earmarked Yes * Foreign vehicle charges: 100 percent earnarked * Local vehicle tax: not earmarked Czech Republic * No earmarking No road fund, but transport fund is * Ministry of Finance decides on yearly budgets under discussion Estonia * Fuel tax: 55 percent (75 percent in future) earmarked No road fund, but it is in discussion Hungary * No, considering earmarked transport fund No road fund, abolished in 1998; however the Hungarian Development Bank acts as if a first generation road fund Latvia * Fuel tax: 50 percent earmarked Yes, global best practice * Vehicle tax: 100 percent earmarked Lithuania * Fuel tax: 100 percent earmarked Abolished * Vehicle tax: 100 percent earmarked Poland * Fuel tax: 30 percent earmarkeda Yes, infrastructure fund * International traffic charges: 100 percent earmarked Romania * fuel tax: - 50-75 percent Yes Vehicle sales tax: 100 percentc Slovak Republic * Fuel tax: No earmarking No road fund, abolished in 2000 * Vehicle tax: 70 percent earmarked * The other 30 percent of vehicle tax to municipalities Slovenia * Fuel tax: -25 percent earmarked ("petrol tolar") No road fund, but commercially managed motorway agency and innovative maintenance concession a. Fuel taxis transferred to state budget, but at least 30 percent must be used for roads. b. Roughly 25 percent of the wholesale price. c. Vehicle sales tax is levied at roughly 10 percent of the wholesale price or value at customs. Source: World Bank. - 65 - Expenditure Policies Toward EU Accession Road funds are often criticized for their extrabudgetary character. However, not all road funds are beyond the control of the ministries of finance and the legislative branch. Even in cases where a road fund behaved as a typical off-budget institution and became indebted, it was often with the "blessing" (if not the encouragement) of the finance ministries. Typical examples are the Slovak and the Hungarian road funds. To simplify both cases can be characterized by the overly ambitious political will to build motorways without full attention to the cost incurred. In Hungary, the approach intended when the road fund was initially abolished was to operate through the budget in a transparent and competitive manner. Unfortunately, the following government reverted to less transparent arrangements than those of the previous road fund. The Hungarian Development Bank (MDF) has become the primary agent of the road development program. It has been exempted from prudent banking principles. Oversight by the government, the road experts, or the public has not been enforced. As a result, its indebtedness brought about by large-scale road investments is huge. At the same time, results remain to be seen in terms of costs, length, or schedule. MDF, for instance, managed to escape public procurement rules, and investment contracts have been granted through nontransparent negotiations, inevitably leading to higher construction costs. Encouragingly, however, following recent elections, the incoming coalition has stated its intent to put an end to this unfortunate episode. On the other hand, a positive example for the institutional solution is the Latvian road fund. This is considered by the international road community as globally best practice. The Latvian model is exemplary and is currently one of the very few in the CEECs that can grow into a much more user-fee-related commercial road management as foreseen in the EU White Paper on fair transport pricing (box 3.4). Despite the good structure, however, the road expenditure in Latvia is not adequate to fund all the needs inherited from the past and in addition created by the European integration. Box 3.4 Key Features of the Latvian Road Fund . Revenue comes from vehicle license fees and a levy added to the price of fuel. The Ministry of Fmance made sure that the levy remained budget neutral (if the road fund levy was increased by 2 cents/liter, the general tax on fuel was increased by 2 cents/liter). General taxes therefore increased, rather than being diverted to the road fund. . The fund has a broad-based oversight board with advisory status. It has 13 members and a chairman (the Minister): 6 represent central government, 2 represent local government, and 6 civil society. The central government members are ex officio, while the rest are nominated by the organizations they represent . The road fund finances both national and municipal roads (about 30 percent to the municipalities). There are transparent and well thought out formulas for dividing the funds between different municipalities (m many respects better than U.S. Federal Highways Trust Fund formulas). . The fund is being rigorously audited. Extensive checks are in place to make sure that the funds are being spent on an approved program and that the work has been carried out according to the specifications. -66 - Successfully Integrating with the European Union Table 3.17 summarizes the arguments for and against road funds. It must be noted that the weaknesses usually occur in the case of the "first generation" road fumds and the strengths are characteristic mostly of the "second generation" road funds (for example, the Latvian Road Fund). Table 3.17 Road Fund SWOT Strengths Weaknesses * Recognized utility function through road fee / * Can be captured by "local powers." road user charges. Users' ownership and * Limits the government's fiscal re-distribution oversight, transparency and accountability, less function and can limit expenditure management. corruption. Multi-annual predictability, multi- * May "over-bozTow" if it loses its year investment planning. Ma "oe-row if t lss is creditworthiness, eventually the burden falls back * Can borrow directly agamist future revenues. on the government, which will have to bail it out. * Cannot be easily mfluenced by political * Difficult to design its re-allocation functions to pressure and thus can stick to economic make them flexible, but not abusive to regional analysis based planning and investments, and local governments if all road user charges go * Prudent financial management, procurement, to the road fund. and road expertise. * Road management and fund management can have contradictory agendas. Opportunities Threats * Can grow into a national road utility that can be * Not appropriately established road fund can be a commercially managed. target of political and public dissatisfaction. * Offers more opportunities for publhc-private * The off-budget status may narrow/limit the partnership and eventually for the pnvate mterest of the government to provide for roads as sector. part of public goods. * With a well designed road fund accelerated * Lack of proper over-sight can lead to road development can be implemented with mismanagement of the road fund and trigger loss lower costs than with central budgeting due to of confidence in the road sector specialists. the risk factor attached to multi-year * Poor road fund management can have a backlash predictability. and can brmg the sector back to central plannmng solutions. * National roads have too much emphasis at the cost of regional, rural, and urban roads. Managing Road Programs Progress has also been made in reforming the road administrations (introduction of a road data bank, pavement and bridge management systems, HDM, etc.), improving their road statistics (again, a number of traffic censuses have been funded by the EU), and converting the road agencies from vertically integrated construction firms with sector management functions to road managers concentrating on contract management and not on implementation. This latter reform started well before the politico-economic changes. Several CEECs had a well- developed construction sector as early as the 1980s, although contractors were still state owned. In Hungary, for example, the privatization of the construction sector took place before or around late 1980s early 1990s. By now, all CEECs have institutionalized the principle of contracting out road works. Most of the countries have done so with all types of works, including routine maintenance. - 67 - Expenditure Policies Toward EU Accession Not enough progress has as yet been made in making the road administrations leaner; that is, concentrating only on the national roads while regional and local authorities could take over the lower category roads. This delay is due to the lower pace of fiscal and institutional decentralization to the regions and municipalities partly to delays in the reclassification of roads, and mostly to the unsolved financing issues. Concluding Remarks While it is true that potential investors might be deterred by the poor road accessibility, the CEECs increasingly recognize that the motorway construction program has to be sustainable. When the construction and financing of the motorway section is being planned, its socioeconomic impact should be compared to investing in the equivalent amount in the preservation and improvement of existing road assets or investing in transport services, or education or health. Road investments, whether to preserve existing road assets (deferred maintenance and routine and periodic maintenance) or to add capacity to the basic network and to expand the motorway network, need to compete for limited funds, with maintenance having the highest priority and new construction having the lowest. The limited budget resources should therefore be allocated to the preservation of the existing assets, and appropriate business conditions should be created for public-private partnership. CONCLUSIONS As the burden of the legacy of the early transition period is reduced, governments around the regions are being faced with new expenditure demands - generally legitimate but often overwhelming-to support their countries' integration into the European Union, and beyond, into the global economy. Many of these demands relate to the need to upgrade skills, environmental standards, and transport networks. Clearly not all can be accommodated within the fiscal frameworks of the PEPs. Governments will need to be all the more vigilant to see that some of these demands come supported by financing on sometimes very attractive terms. Even in those cases, the issue remains to ensure that higher priority claims on public resources are not being displaced and that the programs being financed are sustainable beyond the initial investment. In the education sector, we highlighted the possibility of taking advantage of the ongoing demographic trough (a) to redirect resources from lower to higher levels of education (including by downsizing staff and facilities in many primary and secondary education programs), (b) to focus secondary education on teaching more generic skills for a few broad families of occupational specializations rather than highly specific skills for a large number of narrow occupations, and (c) to expand quality tertiary education, perhaps by concentrating state budget financing on either the most capable institutions or the most capable students, and encouraging accredited private institutions and tuition payments to look after the rest of the sector. As to environmnental protection, the undoubted benefits it will bring will come at a cost. Whether this cost is worth paying for will need to be established on a case-by-case basis. Here we highlighted a number of ways in which the trade-off could be improved by adopting least- - 68 - Successfully Integrating with the European Union cost investment solutions (especially for energy-related investments); opening up procurement to international tender; and phasing compliance measures over time, starting with those that will bring the most immediate benefits to the populations and scheduling the more expensive items later in time. As to transport netwvorks, we highlighted the importance of the following actions: (a) reviewing the road user charges system (considering tax and fee increases, as well as social marginal cost-based road pricing); (b) improving efficiency in spending and in road works (contracting out, many performance based contracts); (c) further reforming road administrations (quasi-utility functions, independent auditing, staff development); (d) basing investment decisions on economnic criteria (maintenance should receive first priority); and (e) reviewing, and if need be reclassifying, the road network. - 69 - 4. ENSURING SOCIAL PROTECTION Alongside the broad consensus that exists in favor of market economics and integration with Europe and the rest of the world, there remains a widespread demand for governments to protect people against the major areas of vulnerability. Health and old age are two particularly sensitive areas. Since early in the transition, however, governments have had to contend with the rising cost of extending such social protection. In a bid to curb this cost, one country after another has revisited the old Bismarckian model of social policy and tried to define a new role and new models for the concept of social insurance that lies at its heart. Here we will discuss the following areas: (a) pensions, where a reassertion of the insurance principle has paved the way for restoring the financial sustainability of public pension schemes, and in some countries for diversifying the forms of old-age coverage into multi-pillar systems; and (b) health, where the focus of government action has generally shifted from providing health care to providing health protection. As is made clear, progress is far more advanced in the area of public pension systems, with patterns of best practices now emerging clearly among CEECs (including progress in comparison with EU members). Health care and insurance reform is as yet at an earlier stage, with most countries still experimenting with various innovations, and the conclusions drawn will accordingly be more tentative. REFORMING PUBLIC PENSION SYSTEMS Pension systems came under serious strain during the 1990s, and, despite various attempts at reform, pension schemes were still displaying deficits in most CEEC countries at the end of the decade (table 4.1). The causes of these deficits varied between two groups of countries: In one group (including Bulgaria, Estonia, Latvia, Hungary, and Poland) pension expenditure had, at the outset, been unsustainably high (table 4.2). These countries undertook systemic reforms and, as a result, are now experiencing transition deficits approaching long- term sustainability. In the other group (including the Czech Republic, the Slovak Republic, and Slovenia) pressures were not felt immediately (or was more successfully contained), but are now increasing.63 63 In the case of Romania, pressures-though always high-were never successfully addressed. Lithuania, having apparently succeeded in stabilizing its first public pillar, is now moving toward establishing a second funded pillar. - 71 - Expenditure Policies Toward EU Accession Table 4.1 Deficit/Surplus of Public Pension Systems in the CEECs, 1991-2000 (percent of GDP) 1991 1993 1995 1996 1997 1998 1999 2000 Bulgaria -0.4 -0.2 0.0 0.4 -0.2 -0.5 -1.4 Czech Republic 0.6 0.7 0.3 -0.5 -0.7 -1.0 -0.9 Estonia 0.3 0.0 -0.4 -0.3 0.2 -1.7 n.a. Hungarya 0.5 -0.2 -0.1 0.0 -0.5 -1.0 -0.6 Latvia -2.4 0.9 -0.5 0.4 -0.5 -1.2 -0.6 Lithuania 1.4 -0.2 0.0 0.2 0.1 -0.7 -0.2 Polandb -4.3 -3.9 -3.2 -3.9 -3.3 -3.9 -3.7 Poland -2.2 -1.5 -0.9 -1.6 -1.2 -1.9 -1.7 Romania -0.8 -1.1 -1.3 -1.6 -1.2 -0.8 Slovak Republic 0.2 -0.1 -0.3 -0.8 -0.2 Slovenia 1.6 -1.6 -3.1 -3.4 -3.8 -4.2 -4.0 a. Includes underage disability pensions of the Health Insurance Fund. b. Includes Farmers' Pension System. Table 4.2 Expenditures of Public Pension Expenditures in the CEECs, 1991-2000 (percent of GDP) 1991 1993 1995 1996 1997 1998 1999 2000 Bulgaria 8.4 7.6 6.4 8.5 8.7 9.0 Czech Republic 7.2 7.5 7.8 8.8 9.0 9.4 9.5 Estonia 6.7 7.1 7.8 7.4 7.2 8.6 Hungarya 10.5 9.1 8.5 8.3 8.7 8.9 8.7 Latvia 9.6 10.4 10.6 10.4 10.6 11.5 10.3 Lithuania 6.1 6.2 6.1 6.4 7.0 7.6 7.2 Polandb 12.5 14.5 14.4 14.4 12.8 12.8 12.7 Poland 10A 12.1 12.0 12.1 10.6 10.8 10.6 Romania 6.5 6.4 5.9 6.6 6.8 7.1 Slovak Republic 7.7 7.6 7.5 7.6 7.7 7.7 Slovenia 10.4 14.1 13.9 13.8 13.8 13.9 14.0 a. Includes underage disability pensions of the Health Insurance Fund. b. Includes Farmers' Pension System. Sources of Pressure The CEEC countries inherited pay-as-you-go schemes (PAYG) that were generally mature and enjoyed nearly universal coverage. Benefits were generous compared with international standards (that is, the minimum retirement age and length of service were low relative to the replacement rate).64 In Poland and Romania farmers also enjoyed separate schemes that were largely noncontributory. The maturity and generosity of these systems left them poorly equipped to withstand the financial strains associated with the transformation to a market economy. 64 In most countries, the minimum retirement age and length of service for women was as low as 55 and 25 years, respectively. The net replacement rate was close to 60 percent in most countries, reaching 64 percent in Czechoslovakia and 80 percent in Slovenia in 1989. In addition, the insured benefited from many types of non- contributory periods. - 72 - Ensuring Social Protection Pressures emerged on both the expenditure and the revenue sides as beneficiaries expanded and the contribution base shrank. All in all, the ratio of pensioners to contributors-a major determinant of the sustainability of a pension system-deteriorated markedly in the region (figure 4.1) and became highly skewed compared to the old-age dependency ratio.65 Figure 4.1 Demographic and Pension System Dependency Ratio 120- 100 80 -. 60 40- 20 - 0 - | Pension DR in 1991 a Pension DR in 2000 E Demographic DR in 2000 Note: DR, dependency ratio. For the Czech Republic and the Slovak Republic, the first column refers to 1995; for Hungary, the middle column is for 1995.The demographlc dependency ratio is calculated as the population above 60 years over those withm the 20-60 age range, and the pension system dependency ratio is estimated as the share of contributors over pensioners. Source: World Bank. Pressures on the Expenditure Side In most countries, the expansion in the number of pensioners exceeded the underlying growth in the elderly population as early retirement, occupational privileges, and disability claims surged (figure 4.2). Growth was most dramatic in Romania, Lithuania, and Poland, and was particularly acute in the early transition years. Indeed, many CEEC countries initially turned to the pension scheme as a mechanism to relieve the escalating number of unemployed and ease the social and political costs of restructuring state enterprises. Early retirement provisions became highly permissive, which placed an excessive implicit tax on an additional year of work.66 Faced with these incentives, rational behavior led many workers to opt for early retirement. Early retirement (as well as disability) can impose a heavy financial burden because it causes both an increase in expenditures and a reduction in the labor force and the revenue base supporting beneficiaries. 65 The formula below illustrates the parameters determining the financial sustainability of a PAYG scheme: f P penslonersp Contsi-ton ofte =- S where - represents the replacement rate (or the ratio of pensions over wages). w contnbutors ' w 66 Consisting of the contribution rate plus the loss of a year's benefits. - 73 - Expenditure Policies Toward EU Accession Figure 4.2 Number of Beneficiaries from Public Pension Systems,1991-2000 (1995 = 100) 130 - 120-7 810 - 80 -~~ ~~~~~~~~~~ 1Ol991 a 1995 012000 Source: World Bank. Among CEEC countries, Romania and Bulgaria presently confront the worst dependency ratios, having reached or surpassed 100 percent, a level at which each contributing worker needs to support one beneficiary. These ratios are highly adverse by any international standards.67 In parallel, many countries suffered a dramatic expansion of occupational privileges with access to early retirement, an inheritance from the planned economy era.68 Criteria for granting disability also became more lax in many of the CEEC economies. Workers not eligible for retirement frequently resorted to disability as an alternative means to escape the highly unstable labor market environment. This problem became particularly acute in Poland, where the number of disabled rose by more than 20 percent during the transition decade. Pressures on the Revenue Side Broadly speaking, developments in the contribution base were even more unfavorable than on the expenditure side (figure 4.3). Most countries suffered an important erosion of their tax base, particularly in the early transition years, as the labor force participation shrank and unemployment soared. In Hungary, for example, the labor force participation rate fell from 85 to 76 percent between 1990 and 1994, and unemployment rose to about 10 percent. The composition of the labor force also underwent a dramatic transformation with a growing number of self-employed operating in the informal market and evading the system.69 67 In Romania, for example, new provisions became effective in 1990 permittmg early retirement up to five years with full benefits for workers satisfying the minimum length of service. Thus, the number of retirees surged by more than 30 percent m a single year. In Bulgaria, policies also permitted early retirement to workers laid off by restructuring firms, prompting the number of pensioners to swell in response. Although this policy was swiftly reversed in 1993, slowmg down the inflow of new retirees, it left the pension system saddled with a greatly expanded beneficiary roll. 68 In Bulgaria and Romania, the number of workers eligible for early retirement reached 20 and 30 percent, respectively. In Poland, approximately 24 percent of the workforce enjoyed sectoral privileges, the cost of which was estimated at 12 percentage points of the 45 percent contribution rate (Chlon, Gora, and Rutkowski, 1999, "Shaping Pension Reform in Poland: Security through Diversity", Social Protection Discussion Paper No. 9805, World Bank, Washington, D.C.). 69 In Romania, for example, the self-employed grew from a very low base to more than 20 percent by 2000. The pension scheme became voluntary for the self-employed and most of them preferred not to join. - 74 - Ensurng Social Protection Figure 4.3 Number of Contributors to the Public Pension System, 1991-2000 (1995 = 100) 150- 130- 110 90 - 70- l ~ F iXmri;; 50 0z1991 141995 Ei2000| Source: World Bank. In Hungary, Latvia, and Lithuania the contribution base largely stabilized by the end of the decade as economnic recovery took off. By contrast, the number of contributors continued to fall sharply in Bulgaria and Romania during the second half of the decade, partly a result of deep economic crises and partly because of the expansion of the informal market.70 These two countries had increased contribution rates (table 4.3) to cope with the financial deficits that further compounded labor market costs and incentives to remain in the informal market. A few countries (for example, Poland and Slovenia) were able to maintain a comparatively stable contribution base throughout the decade. In the Czech Republic, the decline in the contribution base was relatively modest by comparison but it has not yet abated, possibly because of growing unemployment during the second half of the decade and the relatively permissive early retirement criteria that prevailed until 2001. 70 In Romania, the drop in the number of insured was more acute than the fall in employment rates. - 75 - Expenditure Policies Toward EU Accession Table 4.3 Basic Characteristics of Mandatory Pension Systems in CEEC Countries following Transition Periods Country Contribution rate, Retirement age, Public pillar Mandatory public/private public/private structure pillar Switching Effectiveness strategya date Bulgaria 27-24/2-5 63/60 DB M <40 2002 Czech Republic 26 62/61-57b DBc Estonia 16/6 63/63b DBc [20021 Hungary 24-22/6-8 62/62 DB M for new 1998 entrants, V for existing affiliates Latvia 25.5-17.5/2-10 62/62 NDC 2001 Lithuania 25 62.5/60 DBc Poland 25.2-7.3 65/60 NDC 30 < V < 50, 1999 M<30 Romania 33 65/60b DB Slovak Republic [28] 60/57-53c DBc Slovenia [24.4] 63-58/61-58b DB Notes: DB, defined benefit NDC, notional defined contribution. a. M and V refer to mandatory and voluntary switching, respectively. b. Early retirement is permitted on a nearly actuarially fair basis, except in the Slovak Republic where penalties on early retirement are less restrictive. c. The system remains highly redistributive. Initial Ad Hoc Responses As the worsening dependency ratio affected the pension system finances, a new trend emerged, compressing the benefit structure in order to contain expenditures. A lower ceiling was set on the maximum benefit (for example, in Hungary and Poland) or a new flat component was explicitly introduced while the earnings-related component diminished in size (for example, in Estonia, Lithuania, and the Czech Republic). In other countries (for example, Romania and Bulgaria), which suffered a more severe shock, the benefit formula underwent further ad hoc changes. To curtail expenditures, price indexation became more frequent compared to the traditional practice of full wage indexation. In some countries indexation fell behind inflation during periods of high financial pressure. Many CEEC countries did not predefine the indexation rule but granted significant discretion to policymakers, helping them maneuver the difficult and unstable finances in the system and control deficits. However, discretion rendered changes in pensions more unpredictable to beneficiaries and more vulnerable to political cycles.7 Curtailing other entitlements -low retirement age and occupational privileges - could have relieved the pressure on the system dependency ratio but this proved a far more difficult task. A few exceptions can be noted. The Czech Republic, for example, eliminated occupational 71 In Estonia, for example, pension indexation in 1999, an election year, was far above the levels recorded in previous years. In Romanua, changes in benefits have also been influenced by political cycles. - 76 - Ensuring Social Protection privileges early on and introduced a modest rise in the retirement age in 1995.72 Lithuania also mandated a progressive increase in the retirement age in 1995. Ongoing Reform Process However, a fundamental re-thinking of the reform process came about as a result of the following two sets of factors: (a) It was realized that aging demographics would exacerbate the strains over coming decades. As figure 4.4 illustrates, the old-age dependency ratio of the population in CEEC countries is projected to rise dramatically from a range of 30-40 percent in the year 2000 to 55-75 percent in 2040 (b) Years of volatility and compression of benefits sapped the public trust in the public pillar in many countries. Figure 4.4 Demographic Dependency Ratio, 1995-2040 (population above 60/population 20-60) 90 80 70 EU 1995 60 50 02000 40 0~~- 2040 30 20 10- 0 Source: World Bank. This erosion in the confidence in public schemes, together with the prospect of further deterioration, contributed to a growing political consensus in favor of undertaking profound reforms. The outcome was a wave of multi-pillar reforms that was launched in the late 1990s starting with Hungary, Latvia, and Poland and followed by Bulgaria and Estonia.73 Table 4.3 summarizes the main features of the reformed and unreformed systems in CEEC countries. 72 The retirement age will gradually rise to 62 for men and 57-61 for women (depending on the number of children) in 2007. 73 In Lithuania, the draft law establishing the mandatory second pillar could be approved in Parliament over the next few months with implementation starting in 2003. In Romania, the draft law on the mandatory funded pillar is still being finalized. - 77 - Expenditure Policies Toward EU Accession The primary objectives of these multi-pillar reforms were to restore financial sustainability while ensuring adequate retirement income, but other goals also ranked high, especially that of improving microeconomic and labor market incentives, which were much distorted by the high payroll taxes, compressed benefit structure, and early retirement criteria of the former pension schemes. Other positive externalities were also expected, such as a more efficient intermediation of national savings and a greater diversification of pension portfolios. Reforming the Public Pillar The subsequent reforms of the public pillar have basically pursued one of the following two approaches. In some countries, while the structure of benefits was changed, the defined benefit (DB) principle was preserved. Typically the new benefit formulas account for the entire contribution history of the worker and decompress pensions in order to tighten links between contributions and benefits. Early retirement has been eliminated or is permitted on a restricted and actuarially fair basis. Occupational privileges have been abolished in most cases.74 A limited number of countries with highly skewed dependency ratio (for example, Bulgaria, Romania, and Hungary) have complemented changes in the benefit structure with an in increase in the minimum retirement age. In a second group of countries (Latvia and Poland), the public pillar has been transformed into a notional defined contribution (NDC) scheme,75 a new PAYG paradigm developed in Europe in the 1990s.76 The NDC scheme retains the PAYG financing structure, but the replacement rate will be determined by the internal rate of return of the PAYG (that is, growth in the covered wage bill and longevity at the time of retirement).77 This construction will tighten links between contributions and benefits, encouraging individuals to remain longer in the labor force. In addition, the NDC will enhance the financial resilience of the PAYG to demographic and labor market factors that cannot be predicted with certainty.78,79 Both approaches (reformed DB and NDC schemes) have enhanced microeconomic incentives while preserving some key social objectives. For example, these systems maintain a minimum pension guarantee targeted to low-income workers and credit for noncontributory periods related to specific social circumstances (for example, care of young children). In some countries 74 In Poland and Bulgaria an innovative approach has been implemented. Separate funded schemes, privately managed, will finance the pensions of workers with a right to early retirement until they attain the statutory retirement age in the PAYG. This framework enhances the transparency of financing these costly provisions. 75 Hungary is also considering transforming the DB structure into an NDC scheme. 76 The NDC models in Latvia and Poland were inspired by the Swedish reform. The Italian system also follows a broad NDC framework but its design is less resilient to economic and demographic changes. 77 Contributions are recorded in individual accounts and a notional interest rate is credited to the notional accumulated balances. The notional interest rate is determined by parameters highly correlated to revenue growth- (for example, growth rate in the covered wage bill). The pension is calculated by multiplying the notional accumulated balances by an annuity factor that reflects life expectancy. 78 Differng from the Latvian and Polish models, the Swedish scheme incorporates an additional short-term stabilizer. 79 By contrast, in the DB structure, unpredicted demographic or economic changes are more likely to demand legislative amendments requiring a long process of political consensus to amend provisions in the scheme and restore financial stability. - 78 - Ensuring Socal Protection these benefits are now covered by the state budget, allowing greater transparency to the financing of special social programns.80 Establishing Mandatory Funded Schemes Beyond reforming their public pillars, a number of CEECs have established second, funded pillars. Although funded pillars present many variants across countries (table 4.3), they also share important common features, including the following: * Assets are to be privately managed and workers are to freely select the asset manager to prevent inappropriate political interference. * In most countries the private sector not only manages pension assets but also administers the related pension accounts. * To minimize private sector risks, there is to be a clear legal separation between the property rights of fund managers and those of affiliates to prevent any siphoning of resources from workers to the fund administrator. * The funded pillars typically follow a defined contribution (DC) structure that is easier to supervise.81 The size of the second pillar as well the switching strategy (that is, the share of workers allowed to enter the new system) have generally been defined in consideration of the "transition costs" that reforms would generate. The latter occur on the public books because the second pillar is financed by diverting a share of contributions from the PAYG, exposing in the process the existing stock of implicit liabilities within the PAYG. Contributions to the second pillar are planned to range from 5 to 10 percent of wages but can start at lower levels in some countries (such as 2 percent in Latvia and Bulgaria) to ease transition costs. Only Estonia will finance a share of the second pillar by raising the contribution rate of switchers ,to the funded pillar. Other countries have felt that social insurance contribution rates were already burdensome enough. For reasons of fiscal stability and intergenerational equity, CEEC countries implementing multi-pillar reforms have spread the transition costs among different generations. Existing pensioners are the generation least affected by the introduction of the second pillar, as their entitlements have been largely protected, excluding changes in benefit indexation. In some countries, older workers will bear some of the adjustment through a reduction in replacement rates and increases in the minimum retirement age. Other transition costs will be financed through a combination of privatization revenues, fiscal adjustment, and borrowing, transferring some of existing pension debt to future generations. 80 Noncontributory periods have generally been much rationalized in CEEC countries, and nonpriority programs (for example, credit for students), have been eliminated. 81 Over the last two decades, DC plans have been growing at a faster pace than DB plans, mostly offered through occupational schemes, in OECD countries. There is also an increasing tendency for new occupational schemes to follow a DC construction. - 79 - Fxpenditure Policies Toward EU Accession Sustainability Prospects The multi-pillar reforms described have considerably enhanced the long-term financial sustainability of pension schemes in most CEEC countries that undertook reforms (table 4.4). As a result of these reforms, the pension schemes in Bulgaria, Estonia, Hungary, and Latvia are better prepared to confront the aging demographics, even though some additional reforms might still be necessary to attain a balanced position in the long term. In Hungary, for example, the reform was able to narrow the long-term deficit (2050) in the public pillar from nearly 6 percent of GDP to only 0.5 percent. Table 4.4 shows that Lithuania's public pillar is also on a sounder financial footing. Table 4.4 Projected Financial Performance of the Public Pillar Following Reforms Approved through End of 2001 (percent of GDP) 2000 2005 2010 2020 2030 2040 2050 Bulgariaa -1.4 -1.3 0.1 0.3 1.4 0.6 0.6 Czech Republic -0.9 -0.9 -1.3 -3.0 -4.7 -7.6 -8.7 Estonia n.a. 0.4 1.6 1.9 2.2 2.8 2.9 Hungaryb -0.4 0.6 0.0 -0.3 0.0 -0.3 -0.4 Lithuania -0.2 0.5 1.3 1.4 0.6 0.1 -0.9 Polandc -2.3 -2.2 -1.2 -2.5 -2.2 -1.8 -2.2 Romania -0.8 -0.7 0.3 1.0 0.6 -0.1 -0.3 Slovak Republic -0.2 -0.4 -0.6 -2.2 Slovenia a. Preliminary. b. Excludes underage disability pensions of the Health Insurance Fund. c. Excludes the cost of the Farmers' Pension Program. In the medium term, some of these countries will continue to register a deficit as contributions are diverted from the PAYG to the funded pillar,82 calling attention to the need for further reform. In Poland, the reforms have diminished deficits in the public pillar from 7.5 percent to 2.2 percent of GDP in 2050, yet disability remains a burden in the public system and the cause of remaining imbalances. In comparison, however, the group of CEEC countries (the Czech Republic, the Slovak Republic, and Slovenia) that adopted a more gradual approach toward pension reform stands on a much weaker financial footing (table 4.4). Pressure to reform in this group of countries was initially less intense since their system dependency ratios did not deteriorate as severely (Figure 4.4). While in other CEEC economies sharper reductions in benefits had served to discredit public schemes, these three countries were thus able continue to index pensions close to wages. Under the circumstances, a political consensus to undertake anything more than parametric reforms has proven evasive. 82 The initial transition costs will be as high as 1.5 percent of GDP in countries with a larger pillar. - 80 - Ensunng Social Protection Faced with a heightened dependence on budgetary transfers, Slovenia approved a reform of the PAYG in 1999,83 and a new package of parametric reforms is being discussed. The Czech Republic also approved a series of parametric reforms during the second half of the last decade, the most important of which is the 1995 Law. Yet there is a growing realization that, left unattended, deficits are set to expand at a rapid pace. If the more limited parametric changes have occasionally succeeded in narrowing financial gaps, then they have fallen short of building sustainable systems. Thus, while the performance of public pillars in most other CEEC countries will improve over the coming years, financial pressures are mounting in the public schemes of this group of countries. As regards Romania, the figures presented in table 4.4 only suggest that financial viability in the public pillar could potentially be restored if the government used the discretion the pension law provides to set a prudent replacement rate and indexation rule. In practice, however, this discretion has already been used to dissipate some of the early benefits of reform,84 putting the country back onto unsustainable grounds. Challenges Ahead Two broad clusters can therefore be identified: those facing the challenge of consolidating recent reforms and those that still need to define a feasible approach to reform. The first group of countries (Bulgaria, Estonia, Latvia, Hungary, and Poland) will have to address three key issues: (a) enhancing coverage and compliance, (b) consolidating the reformed public pillar, and (c) promoting the safe development of private pension plans. The coverage of pension schemes was greatly diminished during the transition decade in most CEEC countries, and these countries must now rebuild much of the lost contribution bases. Expanding coverage and enhancing compliance will improve the finances of the public pension scheme in the short and medium term and might even permit a modest decline in the hefty contribution rate with a positive labor market impact. If coverage is not reestablished, then a second negative outcome will arise in the long term, because many workers could be left without adequate retirement income, facing the risk of poverty during their old age or imposing an additional cost on the public safety net. Within this first group the minimum retirement age, particularly that of women, remains low compared to levels prevalent in OECD countries (closer to 65 years). In addition to gender inequities, the prevailing low retirement age for women imposes a cost burden on the DB-PAYG, especially since life expectancy of women is higher. In the NDC schemes, benefits 83 This reform package inter alia increased the pensionable age for women and established a system of penalties and bonuses for early and delayed retirement, and decreased the accrual rate. This reform was complemented by the introduction of new fiscal incentives in the voluntary pillar. The amendments to the retirement age and the partial cuts in the accrual rate, albeit a positive step, were not sufficient to attam long-term sustainability (table 4.4). 84The projections for Romania assume that the point value is set at 35 percent of the average wage and pensions are indexed to prices not wages. The new public pillar law, however, stipulates a broad range (30-50 percent of the average wage) withm which the replacement rate can vary and defines no principles for determining it The government has not yet defined what the point value will be in the medium term. This loosely defined benefit structure renders the system highly vulnerable to political interference. If indexation of benefits is not de-linked from the point values system, the deficit will grow to reach 5 percent of GDP by 2050. - 81 - Expenditure Policies Toward EU Accession will vary according to the retirement age, and women face the risk of receiving inadequately low pensions during their old age. For this reason, Latvia is eliminating the gender gap and increasing the retirement age to 62 years for both genders. Similar concerns are emerging in Poland, where the retirement age for women is 60 years. Disability insurance remains a potential source of rising expenditures within the public schemes, particularly as older workers might claim this benefit in response to tighter eligibility requirements for old age benefits.85 The matter is especially critical in Poland, which has inherited a large stock of disabled pensioners. Poland also faces the challenge of revising the Farmers' Pension Scheme, a largely noncontributory scheme costing 2 percent of GDP.86 The safe development of private pension plans will also be of paramount importance if the multi-pillar reform is to reach its ultimate goal of providing adequate retirement. The institutional capacity for supervising the newly established pension plans, which will quickly become the most important institutional investors, needs to be strengthened considerably in most CEEC countries. The assets of pension plans in Hungary and Poland will surpass 20 percent of GDP by 2020. In some cases, a guarantee has been built into the mandatory second pillar through a minimum pension or other mechanism.87 Even without an explicit guarantee, the state will still bear an implicit liability in cases of poor performance owing to the pension plans' mandatory nature. Rigorous provisions have been set for the investment regime of these plans, including quantitative restrictions on asset classes (for example, equity, private fixed-income and foreign investment) to reduce portfolio risks given the early development of capital markets in most CEEC countries. By contrast, investment restrictions in EU countries are largely limited to ceilings on a single issuer or an affiliated party. These tight provisions will need to be progressively relaxed as the experience of pension funds increases and the supervisory capacity matures allowing pension funds to benefit from full integration with EU capital markets. An excessively tight regime will constrain diversification and opportunities to attain a higher return. A second group of countries (including Lithuania, the Czech Republic, the Slovak Republic, and Slovenia) is in the midst of approving further reforms or finding a reform path. In Lithuania, the long-term finances of the public pillar appear sound (table 4.4). And the establishment of the proposed mandatory second pillar should move forward, paving the way for greater diversification of retirement income sources and strengthening the connection 85 In Lithuania, for example, the number of old age pensioners stabilized following the rise in the retirement age. In tum, the number of disabled has been rising at a fast pace. 86Various approaches have been considered in the past, including the partition of the system in two. Higher-income farmers would be switched to the contributory scheme available to other workers, and lower-income farmers would remain in the non-contributory one. But the current capacity to assess farmers' income is weak. 87 In some countries (for example, Poland and Hungary), the second pillar also carries a relative rate of return guarantee, an attempt to protect workers against poorly performing funds. The guarantee is financed from a special reserve created by the pension fund manager from his assets. In the case of Hungary, this guarantee is more loosely constructed allowing for some discretion. In addition, there is a guarantee fund in Poland and Hungary in case the asset manager fails to cover this reserve. In Hungary, the guarantee fund will also cover any shortfall in the second pillar benefit below 25 percent of the value of the first pillar pension. These guarantee funds are financed by pension fund administrators. A shortfall of reserves in these funds could potentially lead to a state liability. - 82- Ensunng Social Protection between overall contributions and benefits, given the relatively flat structure of the public pillar. If the size of the funded pillar remains modest, then it might be desirable to further enhance the earnings-related component of the public pillar to attain a closer link between contributions and benefits. In the Czech Republic, the Slovak Republic, and Slovenia, the quest for a more sustainable reform has proven difficult. New parametric reforms are being discussed in the Slovak Republic and Slovenia, and the pension reform debate is also intensifying in the Czech Republic. The new wave of reforms should seek to restore long-term sustainability, not only to temporarily contain deficits. Continued small parametric reforms will eventually erode trust in the public pillar. The Slovak Republic and the Czech Republic are also seeking to further promote the voluntary pillar, inter alia through occupational schemes, to compensate for the foreseen cut in the public pillar replacement rate. But, the voluntary pillar might not reach the desired coverage of younger workers. Policymakers might also want to explore the diversion of a small share of contributions to the privately funded system as an additional option to enhance overal risk diversification within the pension system. Romania is a case apart. As noted above, its pension system experienced the most severe stress during the transition decade, with the system dependency ratio rising to 95 percent by 2000. A new public pillar law became effective in mid-2001 as part of a broader effort to establish a multi-pillar pension system. The public pillar law introduced some positive steps but retained excessive discretion in the determination of benefits, which could jeopardize its financial sustainability. Moreover, the macroeconomic framework is not yet stable, which could impair the growth of safe financial instruments and the government's capacity to finance transition costs in the public pillar. The elimination of excessive discretion in the public pillar law and identification of a viable replacement rate, increased efforts to improve compliance, and a stronger macroeconomic framework are essential pre-conditions for the successful introduction of a mandatory funded pillar. The country now needs to accelerate the increase in the retirement age and eliminate the retirement age gender gap to further dent the excessive dependency ratio and possibly permit a small decline in social insurance contribution rates, which are the highest among CEEC countries placing Romania at a serious competitive disadvantage. Concluding Remarks The transition decade imposed a severe shock in the PAYG schemes inherited from the centrally planned era. But many of the candidates have emerged with systems that are better placed to face the challenge of the aging demographics, and those that are turning their attention to reform now have a rich body of experience from which to draw. Indeed, many of these systems are now on a stronger long-term financial footing than pension schemes in many EU countries that have not yet found a sustainable reform path. As table 4.5 illustrates, public pension expenditures will rise by more than four percentage points of GDP in at least nine of the EU-15 countries over the next four decades, and in three of them the rise will surpass 6 percent of GDP. - 83 - Expenditure Policies Toward EU Accession Table 4.5 Public Pension Expenditures in EU Member Countries: 2000-50 (percent of GDP) Peaka 2000 2010 2020 2030 2040 2050 change Belgium 10.0 9.9 11.4 13.3 13.7 13.3 3.7 Denmarkb 10.5 12.5 13.8 14.5 14.0 13.3 4.1 Germanyc 11.8 11.2 12.6 15.5 16.6 16.9 5.0 Greece 12.6 12.6 15.4 19.6 23.8 24.8 12.2 Spain 9.4 8.9 9.9 12.6 16.0 17.3 7.9 France 12.1 13.1 15.0 16.0 15.8 4.0 Ireland d 4.6 5.0 6.7 7.6 8.3 9.0 4.4 Italy 13.8 13.9 14.8 15.7 15.7 14.1 2.1 Luxembourg 7.4 7.5 8.2 9.2 9.5 9.3 2.2 Netherlands 7.9 9.1 11.1 13.1 14.1 13.6 6.2 Austria 14.5 14.9 16.0 18.1 18.3 17.0 4.2 Portugal 9.8 11.8 13.1 13.6 13.8 13.2 4.1 Finland 11.3 11.6 12.9 14.9 16.0 15.9 4.7 Sweden 9.0 9.6 10.7 11.4 11.4 10.7 2.6 United Kingdom 5.5 5.1 4.9 5.2 5.0 4.4 -1.1 EU-15 10.4 10.4 11.5 13.0 13.6 13.3 3.2 a. Calculated from five-year interval data. Includes all replacement revenues to people over age 55, before taxes. b. For Denmark, the results indude the semi-funded labor market pension (ATP). If the AT? were excluded, the peak increase would be 2.7 percent of GDP. c. Germany's projections do not incorporate the impact of the 1991 reform. d. Results for Ireland are expressed as a percentage of GNP. Source: Oksanen, "Building a Sustainable Pension System", Paper presented at the Conference on Building a Sustainable Pension System, Prague, November 2001. REFORMING HEALTH CARE Following a decade of escalation in the cost of health care, almost all the CEECs have now transformed the basic setup of their health care system in a bid to raise additional resources for the sector and to curb costs through efficiency gains. These objectives would be achieved through the separation of financing the provision of health care, the introduction of social health insurance, the establishment of new provider payment mechanisms, decentralization, and competition among providers and greater choice for patients. Although in many countries the reforms are too recent to judge, a few tentative observations may be made: * While countries have been experimenting with various provider payment arrangements there is little sign as yet that health costs would have been successfully contained. Difficulties in collecting social insurance contributions sometimes exacerbate the resulting financial strains. - 84 - Ensuring Social Protection * What has changed since the mid-1990s is that the further rise in health costs has generally been associated with higher out-of-pocket payments on the part of patients rather than with additional fiscal resources. * These developments have contributed to maintaining or even improving the health status of the population. Wide variations persist, however, across countries. Overall Trends While the CEECs are not the only countries that have experienced an escalation in health costs during the last decade, the increase has been particularly pronounced among the CEECs. Measured in dollar terms (valued at purchasing terms), health expenditure per capita surged by over 50 percent during the past decade (figure 4.5).88 This compares with a 25 percent increase on average 89 among the cohesion countries. The only countries to escape this upward trend were Bulgaria, Lithuania, and Romania (in part because they could not afford it). While total health expenditure remained, on average, slightly below the levels observed among cohesion countries relative to GDP (6.9 percent of GDP compared to about 7.5 percent at the end of the 1990s, respectively), health expenditures per capita were rapidly catching up in absolute levels with those in the cohesion countries in at least two of the CEECs (the Czech Republic and Slovenia) (table 4.6). Figure 4.5 Health Care Expenditure in the CEECs, 1990-99 7.0 - 600 6.5 - t60- M _ 2 30 'S50 ~ ~ ~ ~ ~~~~~~~~0 3.0 - 300 1991 1992 1993 1994 1995 1996 1997 1998 1999 sqpublic expenditure to GDP private expenditure to GDP -U per capita Source: World Bank World Development Indicators, (Health Expenditure Data). B8 Unweighted average. 89 Unweighted average. -85 - Expenditure Policies Toward EUAccession Table 4.6 Trends in Health Care Expenditures in the CEECs, 1990-99 Health expenditure per capita Total health expenditure as (US$ at purchasing power parity, 1999) percent of GDP Percent Percent Total public private 1990 1999 Bulgaria 211.9 95.1 4.9 4.1 4.1 Czech Republic 954 91.6 8.5 5.4 7.1 Estonia 571 79.7 20.3 1.9 6.4 Hungary 756 76.5 23.5 5.7 6.8 Latvia 383 60.6 39.4 2.5 6.6 Lithuania 424 - 76.2 23.8 3 6.3 Poland 522 75.4 24.6 4.6 6.2 Romania 337 71.7 28.3 2.7 5.3 Slovak Republic 826 79.2 20.8 5.4 7.2 Slovenia 1,188 88.0 12.0 5.6 7.5 Memo item: Cohesion Countries 1,235 67 33 6.3 7.4 Sources: World Development Indicators, World Bank (Health Expenditure data). While costs may have increased, by and large the health status of the CEEC population has been maintained or has improved to levels that do not compare unfavorably with those within the EU (table 4.7). Life expectancy has improved-to between 70.2 and 75.8 years-and so has infant mortality. Infant mortality remains, however, very high in countries where health spending is lowest (Bulgaria, Latvia, and Romania). The same countries also show high standardized death rates from cardiovascular and cerebro-vascular ailments. Looking beyond strictly clinical factors, the incidence of lifestyle-related death patterns remains very high among a broader group of countries (including also Estonia and Hungary).90 90 Related to nutritional habits, physical activity, and smoking or heavy alcohol consumption together with the prevalence of such risk factors as elevated blood pressure, high serum cholesterol, or overweight. - 86 - Table 4.7 Comparison of Health Status Indicators between the CEECs and the EU Life expectancy Infant mortality Standard death rates (per 100,000) (years) rate (per 1,000) Selected Selected Ichemic alcohol smoking Circulatory Cerebro- heart Tuber- related related All 1990 1999 1990 1999 system vascular diseases culosis causes causes causes Bulgaria 71.4 71.7 14.8 14.6 721.6 198.7 206.9 3.2 101.6 457.5 1134.9 Czech Republic 71.7 74.9 10.8 4.6 487.2 135.9 199.0 1.1 84.6 408.5 911.0 Estonia 69.5 71.1 12.4 9.5 570.7 165.4 337.6 9.8 173.8 566.4 1095.1 Hungary 69.3 70.8 14.8 8.4 588.5 153.3 254.1 4.3 180.0 539.3 1192.6 Latvia 69.3 70.2 13.7 11.3 604.2 217.9 326.7 12.6 186.2 608.7 1150.9 Lithuania 7i.3 72.4 10.3 8.7 509.0 119.2 307.1 10.0 166.6 503.2 980.0 Poland 70.9 73.2 19.3 8.9 469.0 107.8 147.9 2.9 97.5 337.8 993.4 Romania 69.7 70.6 26.9 18.6 719.8 231.0 249.7 9.7 120.9 558.6 1160.9 Slovak Republic 70.9 72.0 12.0 8.6 541.0 86.8 275.6 1.0 96.9 444.4 1057.8 Slovenia 73.3 75.8 8.4 4.6 330.9 92.0 114.0 1.3 120.6 293.0 832.7 EU average 75.8 78.1 7.7 5.2 269.8 68.6 109.1 0.9 81.8 247.9 690.1 Note: All country data are for the year 1999 while the EU average is for the year 1997. Source: WHO, 2001, Health for All Database. Expenditure Policies Toward EU Accession Many factors have contributed to raising health costs. Some of these factors are exogenous to the sector including a slow aging of the population and the impact of the decompression of earnings that accompanied the transition on medical personnel. In some countries, the trend may also be the impact of a general modernization in medical equipment (investment decisions that may have been made in an uncoordinated fashion by decentralized hospitals or local governments, but may at least have resulted in improved health care). As the discussion below points up, other influences can be traced to persistent, and sometimes growing, imbalances within the sector. Initially, public resources underwrote these extra costs. As the decade progressed, however, the burden of footing the bill increasingly fell on the patients. When the decade drew to a close, out-of-pocket payments,-relatively negligible in the aggregate at the beginning of the transition, had grown to cover (on average) as much as 25 percent of the health care bill (figure 4.5). This percentage was higher than in Belgium, for example, where only 17 percent of total health expenditure came from copayments (of the 7.6 percent of GDP that the country spent on health care in 1997, another 38 percent came from general taxation, 36 percent from social insurance, and 9 percent from other sources). Within the public sector itself, the structure of health financing had by then fundamentally changed. Reflecting the general introduction of social health insurance schemes, social insurance contributions -nonexistent at the beginning of the decade - are taking up the lion' share of the funding in all CEECs except Latvia (table 4.8). Table 4.8 Share of Social Health Insurance and Taxation in Public Spending on Health Year Social insurance Other public resources Bulgaria 2001 40.1 59.9 Czech Republic 1999 89.5 10.5 Estonia 1999 88.3 11.7 Hungary 1996 85.6 14.4 Slovak Republic 1999 74.8 25.2 Slovenia 1999 96.2 3.8 Latvia 2001 0 100.0 Lithuania 200C 74.7 25.3 Poland 1999 82.8 17.2 Romania 2001 87.3 12.7 Sources: World Bank public expenditure reviews; national health accounts (Estonia); State Patient Fund (Lithuania); health sector studies (Slovak Republic, Poland, Romania). Legacy and Change To understand the changes that health systems experienced it is useful first to understand their previous position. Following the Siemaszko model of health financing and organization, the transition economies of Central and Eastem Europe developed publicly funded and operated health systems after World War II. Financed by government revenues, the health care production and delivery system offered universal access and broad coverage. The national budget, either directly through the ministry of health or through other ministries, supported a huge network of state-financed hospitals and clinics. Funds flowed from the - 88 - Ensuring Social Protection central ministries to provincial and municipal governments that, in turn, allocated funds to health providers. By and large, the allocation of funds was made on a historical basis, with some adjustments for inflation and epidemiology, and in some cases the allocation was based on population demographics. Outpatient and primary health services were generally provided in regional and county clinics or health centers at the places of employment. Secondary services were provided mainly in provincial hospitals, while specialized services were provided in university hospitals, medical academies, and science and research institutes. Most health providers were state employees and were paid on a salary basis. While the socialist health systems succeeded in providing risk protection and equity across the population, technical and allocative inefficiencies were commonplace and clinical quality was low, largely owing to the inherent system of incentives and a system of practice in response to these incentives. First, given the nature of the allocation of funds, most health providers had no incentives to develop fiscal and strategic planning functions. The predictability of budgetary allocations undermined the need to improve managerial and organizational capacity, which effectively slowed the process of innovation and the ability to respond to changes. Second, paying health care providers on the basis of salaries undermined the importance of effort and productivity. As a consequence, there was little effort to improve efficiency and the quality of care, as a result of which patients faced erratic service. Third, the reorganization and restructuring of the public sector was slow, as a result of which the public health care system came to be characterized by overstaffing, misallocation of resources, underutilization of capacity in some areas and under-supply in others, and shortages of drugs and medical supplies. The transition brought into the open a widespread dissatisfaction with the health system. Patients were dissatisfied with the quality of care, with restrictions in the choice of provider, and with increasing unofficial payments. Medical personnel were dissatisfied with low compensation levels. Cost inflation, new technology, and rising consumer expectations further exacerbated the situation. The loss of fiscal resources that most CEECs experienced at the beginning of the transition made the entire arrangements increasingly unsustainable. Therefore, in a bid to maintain earlier levels of coverage, governments around the region sought to raise nonbudgetary sources of revenues for the sector, while instigating improvements in the use of existing resources. In doing this, many countries of in the region, invoking the principles of solidarity, moved to the Bismarckian model of predominantly social insurance-based financing. Hungary opened the way in 1990 and the other CEECs followed. By the turn of the millennium, only Latvia still ran a (reformed) single payer system (table 4.9). - 89 - Expenditure Policies Toward EU Accession Table 4.9 Social Insurance Contribution Rate for Health Salaried Self-employed Year (employer (percent of declared Non-active introduced employee) income) population Bulgaria 1999 6 percent of payroll 6 percent General budget transfer. Non-active (3:3) population includes registered unemployed and social assistance Czech 1993 13.5 percent 13.5 percent of 35 Central budget transfer 13.5 percent of 80 Republic (10:3.5) percent of net percent of statutory minimum wage. Non- pretax income active population includes unemployed, pensioners, children and dependants up to 26 years of age, students, women on maternity leave, military, prisoners, and people receiving social welfare Estonia 1992 13 percent 13 percent Central budget transfer. Non-active (13:0) population includes unemployed and military service Hungary 1990 14 percent 14 percent plus Central budget. Per capita amount of transfer (11:3) hypothecated tax is unspecified. Non-active population includes of US$170 per person pensioners, unemployed and sociaUy indigent Slovak 1994 13.7 percent 13.7 percent Central budget. Per capita amount of transfer Republic (10:3.7) specified as the contribution rate applied to 73 percent of the statutory minimum wage. Non- active population includes children, pensioners, persons caring for children or disabled persons, military prisoners, and refugees Slovenia 1993 13.25 percent 13.25 percent Central budget. Non-active population includes Latvia 1998 28.4 percent of 28.4 percent General budget transfer personal income tax Lithuania 1997 3 percent of payroll 3 percent General budget transfer. Non-active tax plus one third of population includes children less than 18 income tax years, students, beneficiaries of social assistance and social insurance cash benefits, and persons with certain illnesses Poland 1999 7.75 percent of 7.75 percent 7.75 percent of gross benefits or central budget taxable personal Non-active population includes registered income unemployed and those on social assistance Romania 1999 14 percent 7 percent 7 percent income tax based on gross benefits. (7:7) Non-active population includes children and young people, the handicapped and war veterans with no income, and dependants of an insured person without their own income Sources: Preker, Jakab and Schneider (2000); Erosion of Financial Protection in Health Systems of ECA Transition Economies (first four columns, all countries except Bulgaria and Lithuania); European Observatory on Health Care Systems. While undertaking reforms, most countries have nonetheless retained the inherited laws and constitutional guarantees to free access, maintaining, in principle at least, universal coverage for health services (though many countries have set eligibility conditions that, in practice, do limit access). Indeed, a common feature in all the 10 CEECs is the - 90 - Ensuring Social Protection comprehensiveness of services covered under social health insurance. Typical exclusions concern only some drugs, elective cosmetic surgery, and most forms of dental care; in some countries, treatment in spas and sanitariums is also excluded. Following reforms in maniagement and in the organization of the health sector in particular, and shifts in political and economic ideology in general, significant changes occurred - and are still occurring - in the system of delivery of health care services and the market for providers. With decentralization and the introduction of new physician payment mechanisms, a new class of publicly financed private providers developed. Private practice by physicians (known even during the days of socialism but not widespread) witnessed an exponential growth. As a result, the provision of health services took on a very complex form, with a wide variety of providers, distinguished both by sources of financing and by their ownership status, operating in a health system characterized by significant interactions between the public and private sectors. Significant changes have taken place also in the ways in which providers are paid. Though varying in content and scale, the general direction of these changes has been toward performance-based payment systems and away from the traditional budgetary allocations and salary-based compensation. The greatest change has come about in reirnbursing physicians for primary care, with all countries shifting toward capitation-based payments for physicians (table 4.10). Most countries are using fee-for-service payments for outpatient specialist care, although some-such as Bulgaria and the Slovak Republic-still pay specialists on a salary basis. As regards paying for inpatient services, some countries -such as Bulgaria, the Czech Republic, and Romania -have retained the traditional budgetary system. Estonia, Slovenia, and Latvia compensate hospitals on a per day basis, while in Poland hospitals are paid a lump sum per admission, with the payment varying by department. Table 4.10 Paying Health Care Providers Primary care Outpatient specialist care Inpatient care Bulgaria Capitation payment Salary (performance pay) Budget Czech Republic Capitation payment Capped fee-for-service Budget Estonia Mix of capitation payment and Capped fee-for-service Per diem payment fee-for-service Hungary Capitation payment Capped fee-for-service DRG (758 categories) (point system) Slovak Republic Capitation payment Salary/fee-for-service Budget Slovenia Capitation payment Salary Per diem payment Latvia Mix of capitation payment and Salary + point system Per diem payment fee-for-service Lithuania Salary/capitation payment Salary/fee-for-service Case-based payments Poland Capitation payment Capped fee-for-service Per admission Romania Mix of capitation payment and Capped fee-for-service Global budget fee-for-service Note. Diagnosis related groups (DRG). Sources: Health Care Systems in Transition (various years), European Observatory on Health Care Systems; World Bank. On the revenue side, contribution rates vary from 7.75 percent of income in Poland to 14 percent in Hungary. The burden is generally shared between the employer and the employee, with the employer paying the bulk of the premium in most countries (table 4.10). Contribution - 91 - Expenditure Policies Toward EU Accession rates for the self-employed are the same in many countries, although most countries have introduced minimum and maximum contributions. The non-active population-the definition of which varies across countries but generally includes pensioners, registered unemployed, the poor, and in some cases children and family dependants -is covered by explicit transfers from the central budget. Fiscal Impacts The various reforms have been successful in at least two of their objectives: maintaining health coverage and raising nonbudgetary resources for the sector (although these resources may not be additional). Many countries, however, are still struggling to secure the expected efficiency gains from the reforms and to put the new arrangements on a firm financial footing. These two outstanding issues will now be examined in turn. Cost Containment Despite the reforms, little progress has been made in correcting the prevailing inefficiencies in the supply of health services and in containing health costs. On the contrary, the huge infrastructure of hospitals, hospital beds, polyclinics, health centers, and health personnel inherited from the communist days continues to drain public resources. In some areas, the imbalances have actually worsened. While all the CEECs succeeded in bringing about significant reductions in the number of hospital beds following the transition, only a few-such as Latvia and Estonia-also succeeded in reducing the number of physicians over time (table 4.11). The demands of maintaining a large physical infrastructure alone require significant outlays of health expenditures, effectively reducing the available resources for drugs, equipment, and consumables. Similarly, many of these countries were unable to reorient service provision from the relatively expensive acute hospital sector to lower levels of care. As a result, hospital admission rates increased in almost all countries while outpatient visits fell substantially in 6 of the 10 countries. Table 4.11 Hospital Beds, Physician, and Utilization Patterns Hospital beds Physicians per Hospital admission Outpatient consultations per 1,000 100,000 rates per 100 per person (percent change, (percent change, (percent change, (percent change, 1990-99) 1990-99) 1990-99) 1990-99) Bulgaria -23.3 8.52 -16.82 -18.18 Czech Republic -22.44 11.81 7.07 4.32 Estonia -37.94 -12.03 6.36 -20.25 Hungary -14.74 12.62 16.47 25.81 Slovak Republic -9.50 18.46 17.99 20.37 Slovema -8.11 4.88 5.99 13.23 Latvia -36.75 -23.84 -1.69 -39.51 Lithuania -24.60 -1.75 31.42 -15.38 Poland -22.73 5.61 30.19 -10.00 Romania -18.05 6.11 3.08 -22.89 Source: World Bank Part of this syndrome can be attributed to persistent organizational inflexibility, even after the reforms, but it also reflects the uneven progress in actually separating health care - 92 - .Ensunng Social Protecton financing and provision and in establishing hard budget constraints and competition among providers and the difficulties countries have encountered in designing. incentive-compatible provider payment systems as part of the financing reforms. Separation of Financing and Provision. The central thrust of the reforms was to create quasi-market situations in which decentralized health care providers compete to attract patients and insurance companies to purchase their services. In practice, however, the reality is that, in most countries under review, the health insurance companies still perform little in the way of insurance functions, be they claims adjustments or quality control-not to mention the setting of premiums and the management of risk exposures (as indemnity insurance would do). Until recently, in the Slovak Republic, for example, insurance companies largely served as passive conduits to channel insurance contributions to hospitals, based on input-driven, ex ante budgets rather than on competition for actual services. These insurance companies are now shifting toward a more active "structural contracting" for bulk purchases from hospitals. Furthermore, despite the presence of multiple health insurance companies in some countries, there is often neither scope for meaningful competition (see below) nor incentives for insurance companies to economize or to be selective in their purchase of health services. Similarly, insurance contribution rates, health service prices, and benefit packages are all strictly and uniformly regulated. Under the circumstances, there might be very few positive benefits to offset the fragmentation and administrative inefficiencies introduced in the contribution collection processes. There also continues to be considerable ambiguity in many countries regarding capital and investment funding, as a result of which health facilities keep postponing repair and maintenance until they become critical (and more expensive). Clear capital and investment policies need to be laid out in order to ensure a sustainable long-run capital investment and replacement regime. Competition among Providers. The scope is for meaningful competition among providers varies across countries and health practices. While competitive private practices are now the norm among CEECs, in countries like the Slovak Republic, for example, hospitals have remained, for all practical purposes, essentially government departments with limited accountability and incentives to compete (this should change under the recently approved transfer of ownership program). Even if such incentives existed, hospitals might not react to them in the absence of hard budget constraints. In contrast, one reason why the Czech Republic has made good progress with the new system, is that it succeeded early on in incorporating, privatizing, or otherwise decentralizing hospitals away from the former ministry of health. Provider Payment Mechanisms. While the reform concept of mimicking competitive market arrangements has considerable intellectual appeal, defining the provider payment formulas that would make it work has often proved to be a challenge. Some of the dilemmas are as follows. For outpatient care, physicians in most countries are paid on the basis of a fixed capitation fee per enrollee for all outpatient services, regardless of the extent and nature of the treatment sought. Physicians participating in this scheme bear most of the risk of treating a patient, and to this extent are likely to be conservative in the amount of health care they provide. Physicians on capitation payment are thus more likely to err on the side of over - 93 - Expenditure Policies Toward EU Accession referring patients to specialists. This often becomes a beneficial arrangement for the physician and for the patients, who prefer specialist care even if the primary care physician can equally well provide the required treatment. However, this increases the costs to the health system since specialist care is relatively more expensive. To compound the problem, apparently more market-driven arrangements can open the way to fee escalation. This has been the experience with the recent provider payment reforms in Bulgaria (under the guise of performance-based pay). It is to avoid such situations that most CEECs now rely on regulated fee-for-service arrangements for specialists. Similarly, prevailing systems of reimbursing hospitals often leave loopholes where providers can escape pressures to contain costs. Paying separately for each service rendered encourages increased use of interventions whose costs can be reimbursed. Paying per bed-day encourages longer hospital stays because the days cost less toward the end of a stay. Reimbursing hospitals for all inputs used leaves the facility managers with no incentives to contain costs. In an effort to reach more satisfactory outcomes, a number of countries are now experimenting with some success with mixed payments systems (table 4.9) combining multiple rewards and penalties to deal with the multiple forms and opportunities for performances and abuses. This is the case with Romania, Estonia, and Latvia for general practitioners and with Latvia, Lithuania, and the Slovak Republic for specialists. Hungary and Lithuania have also moved toward case-mix based payment systems for inpatient services. In the hospital sector, the prospective payment methods based on diagnosis related groups (DRGs) also show a growing appeal. Under such a mechanism, hospitals are paid based on an ex ante estimate of the "reasonable" costs that could be expected given the patient's condition upon presentation, rather than as reimbursed for "actual costs incurred." Successfully implemented in Austria, this system seems to reduce the temptation for hospitals to turn expensive patients away if the hospitals were assured of being paid relatively more (as with a fixed budget), while providing a strong inducement to control costs. Hungary has been pioneering this approach, and some features of it are being incorporated in the provider payment system of Bulgaria. The Czech Republic regards its current fixed-budget system as a transitory step toward a DRG-type of arrangement that it is currently testing with a group of hospitals. Resource Mobilization A number of countries have faced difficulties in collecting health insurance contributions. To some extent, this is part of a more general phenomenon of noncompliance that affects not only taxes and social insurance, but also other dues to utilities, suppliers, and banks. Other factors, more specific to health insurance, have also played a role. First, contribution rates are typically high in comparison with those applicable in EU countries, reaching 13-14 percent of salaries in the largest number of countries (table 4.8). Such rates are particularly onerous given that dependants are often not covered by direct insurance payments. - 94 - Ensuring Social Protection In Austria, in contrast, contribution rates (2000) vary from 6.4 percent for farmers to 9.1 percent for the self-employed and internationals. White-collar employees contribute 6.9 percent of wages (employers pay 51 percent), blue-collar employees contribute 7.9 percent (employers pay 50 percent), and civil servants pay 7.1 percent (employers pay 45 percent). In Belgium, the contribution rate (1997) is 7.35 percent for salaried persons, with the employer paying 52 percent, and 3.2 percent of gross earnings for the self-employed.91 Furthermore, incentives for voluntary compliance are decidedly weakened when, as is generally the case, delinquency has no bearing on coverage. This blurs the distinction between a social insurance contribution and a tax. Thus, it can be hard to avoid "free rider" behaviors on the part of potential contributors (or their employers). Finally, in a number of countries, the responsibility for collecting health contributions is left with the insurance companies themselves. The insurance companies may have little experience in the matter, and little effectiveness in pressing claims perceived to have lower seniority than, for example, taxes or pension contributions. The solution adopted by Poland of assigning a single body to collect both pension and health insurance contributions seems in this respect to have considerable merit. This is also the plan in the Slovak Republic. Illiquidity One symptom of the divergence between revenue and cost containment performances has been the emergence in a number of countries mounting debts in the health systems (box 4.1). Out-of-Pocket Paymnents One symptom of this divergence has been the spread of out-of-pocket payments on the part of patients. As noted in figure 4.5, such payments have emerged as a major source of health care financing. Out-of-pocket payments traditionally take two forms: formal copayments and informal out-of-pocket payments. Payments for pharmaceuticals generally account for the largest part. Although hard numbers are lacking to substantiate this assertion, informal payments may well account for the next largest item. 91 In Italy (1999), 90 percent of the regional tax collected under IRAP (zmposta regionale attiznta sulle produttive) is earmarked for health care. IRAP is levied at the rate of 9.6 percent of the first E20,660 salary for public sector employees and 3.8 percent on the next Euro 56,820, with the employer paying all the tax. For others, a 4.25 percent value-added tax is imposed on companies, corporations, and the self-employed. In addition, a piggyback tax (IRPEF) of 0.5 percent of gross incomes is levied, with 100 percent of the collections under this tax going for health care. - 95 - Expenditure Policies Toward EU Accession Box 4.1 Indebtedness in the Health Systems Despite the increase m spending on health in almost all countries under review, a fundamental challenge that has emerged in most countries is the heavy amount of mdebtedness in the health care system. For example: • In Romarua, late payments by state owned hospitals reached lei 4.6 million in 2001, equivalent to 9.7 percent of total health expenditures. This placed tremendous pressures on the implementation of the 2001 budget, leading to a reduction of expenditures in key services as well as to a rapid build-up of new arrears by hospitals. o In Latvia, debts of public hospitals to suppliers rose by 84 percent between 1998 and 1999, and amounted to 4.4 percent of revenues. Hospital debts increased further in 2000, and payables reached 21.3 million last, equivalent to 11.5 percent of total government expenditure on health in 2000. In response to fiscal constraints, public hospitals have tended to take short-term measures to defer cash expenditures rather than long-term measures to increase efficiency. o In Poland, as a percentage of total public expenditures on health, end-of-year debts fell from 5.4 percent in 1996 to 2.7 percent in 1997, but increased fivefold in 1998 to 13 percent. Eventually, in 1999, the government had to take over all debts of health providers. However, during the last three years and as a consequence of the mcomplete health care reforms introduced in 1999, the arrears and debt of the health providers (read hospitals) have mcreased again to around PLN5.6 billion or around 0.75 percent of GDP. o In Bulgaria, although the last two years have experienced surpluses in the balances of the National Health Insurance Fund, expenditures on health are expected to exceed revenues with effect from 2003, and deficits are projected to reach about 20 percent of revenues of the Fund. * In the Slovak Republic, the receivables of health insurance agencies have grown significantly since 1994, particularly in the area of unpaid contributions. In 1998, almost one-third of projected revenues could not be collected, as a result of which sufficient financial resources were not available in the health care system to support expenditures. As a result, debts of health insurance companies increased from about 16 percent of total public expenditures on health in 1994 to 26 percent in 1998. o In Hungary, following a decrease in the real value of hospital revenue and an increase in the number of cases, a downward spiral in financing inpatient care services resulted, and many hospitals faced a severe financial crisLs and ran up 15-20 percent annual budget deficits. In response to the continuing deficit in the budget of the Health Insurance Fund and prompted by the 1997 revenue crisis of the Fund, the government introduced a hypothecated tax - also known as "health care contribution" - as part of the economic stabilization program. * In Lithuania, the State Patient Fund recorded a deficit of 1.7 percent of total public expenditures on health in 1998, followed by an increase m the deficit rate to 2.9 percent in 1999. Exact figures are not available for 2000, but it is widely believed that the State Patient Fund owes Euro 57.9 million to the drug wholesalers. Informal payments have far-reaching negative implications for health care systems.92 By their very nature, informal payments are unauthorized and contribute to the general environment of corrupt practices and the growth of a parallel health care financing system. Variously referred to as "envelope payments" or "gray payments," informal payments 92 Informal payments, defined as payments in cash or in kind, made by patients or others on behalf of the patients, to public health care providers for health services received or expected to be received, existed even during socialist times. While the state-financed and publicly provided system provided health care free of charge (or for very modest co-payments), it was not uncommon for patients to make direct payments to physicians, nurses, and others connected with the delivery of health services. - 96 - Ensuring Social Protection introduce perverse incentives in the health system and compromise governments' efforts to improve efficiency, accountability, and equity in the public sector. The nontransparent and discretionary nature of informal payments adversely affects equity and access to health care, with the more vulnerable segments of the population having to pay disproportionately large amounts for the health services that should otherwise be available free of charge. To some extent, however, informal payments may be seen merely as a manifestation of the underlying excess demand for (purportedly free) health care. It would be preferable to deal with such an excess demand in an open and transparent fashion, even if it meant shedding the illusion of free health care for all citizens. This would imply a greater use of formal, regulated copayments. In recent years, however, many countries have introduced official copayments, typically applied to dental care covered by health insurance, pharmaceuticals, and medical prosthesis; over time, such official payments have assumed significant proportions. A World Bank survey carried out in Poland in 1998 found, for example, that over 50 percent of all health expenditures on ambulatory care were out-of-pocket, the bulk of these expenditures going to the purchase of pharmaceuticals. Concluding Remarks There is no doubt that many positive achievements have been recorded in the years following the transition in the health reforms in most of the countries being reviewed. If health costs have risen, then so has the health status of the population; and after the mid-1990s, the growth of public spending on health has generally been arrested, albeit at very different levels across countries. Despite sweeping changes to the design of health systems, there is as yet little sign of the expected efficiency gains. This may be because considerably less change has taken place in the way the sector actually operates. In addition, countries are still struggling to define a payment formula to tie their general practitioners, specialists, and hospitals into an appropriate framework of efficiency-enhancing incentives. In this context, a number of recent developments appear to hold the promise of success, including, in particular, the following developments: * Separation of providers from ministries of health (beyond a core number of public health and last resort institutions) and the incorporation of these providers into various forms of ownership * Spread of mixed provider payment formulas and the first experimentation with DRG schemes for the payment of hospitals * Merger of health insurance collection with that of other social security contributions * Regularization of private payments into systems of well-regulated patient copayments. - 97 - 5. AN EXPENDITURE STRATEGY FOR GROWTH AND CONVERGENCE Expenditure policies can make a decisive contribution to the CEECs' objective of growth and convergence with the EU. From what we have discussed, the outline of what such an expenditure strategy might involve begins to emerge. In keeping with the trends observed within the EU (including within the cohesion group) as well as the OECD at large, a first element of such strategy would be to curb aggregate expenditure levels, as most countries have indeed suggested in their PEPs. A first rationale for such an approach is to leave more resources in the hands of a more efficient private sector and thus achieve a more rapid factor productivity growth. With equal importance, such a strategy would create the fiscal space needed to reduce the burden of labor taxation, thereby activating resources (that is, labor) that are currently sitting idle. At present, the burden of labor taxation in the CEECs (at about 75 percent of wages) exceeds even the high levels that saddle EU labor (at an average rate of 53 percent) and is even further out of line with the levels observed in the broader OECD).93 It is hard to see how the CEECs could to live up to the aspirations of the "Lisbon Strategy" they have recently espoused of turning the enlarged EU into "the most competitive economy in the world with social cohesion and full employment of the 70 percent of the adult population" without bringing these tax rates down. A primary tool for curbing expenditure should be to rein in subsidies and transfers. Currently, transfers make up as much as 60 percent of total primary expenditure in the Czech Republic and the Slovak Republic, as compared to 40 percent of primary expenditure in the EU. Even at other end of the range, Lithuania still devotes relatively more of its expenditure envelope to subsidies and transfers (38 percent) than the average cohesion country (35 percent). Indeed, despite years of progress in tightening budget constraints, considerable scope remains in a number of countries to cut back on enterprise subsidies. This is the case of Hungary, Latvia, and the Slovak Republic, where enterprise subsidies (current and capital) account for about 7 percent of GDP. But even Bulgaria, the Czech Republic, Estonia, Romania, and Slovenia still spend more than twice as much relative to GDP (about 4 percent of GDP) as the average EU country on subsidizing various categories of enterprises. As noted, part of the answer lies in completing bank restructuring and expediting large- scale enterprise privatization. In some cases (for example, coal mines) the initial restructuring costs may be high. Provided the restructuring strategies are well tailored to the ultimate privatization of the enterprises, such up front costs are likely to be worthwhile for the sake of the longer term fiscal benefits that will arise from getting ailing firms off of the public books. 93 See Funck, Bernard, and Lodovico Pizzati, 2002, Labor, Employment, and Social Policies in the EU Enlargement Process, World Bank, Washington, D.C. - 99- Expenditure Policies Toward EU Accession A second source of fiscal relief will come from the implementation of the acquis communautaire in sectors such as power, gas, and transport, because it protects government from the ever-present temptation of providing state aid. Similarly, much more could be done to curtail social transfers. In Bulgaria, the Czech Republic, and Hungary, social transfers exceed the levels observed in the cohesion countries. In Slovenia and Poland they are even higher (relative to GDP) than among EU members. As was noted, some of the CEECs have begun to take the measure of the situation and are acting, apparently successfully, to address it (witness the 2-3 percentage points of GDP reduction in Poland and Hungary between 1995 and 2000). But in the absence of systemic action, the upward trend appears inexorable in the Czech Republic (where spending on social transfers has risen by more than 3 percentage points of GDP between 1995 and 2000), and the Slovak Republic may well soon follow suit. The time has therefore come, even for those countries that avoided pension system crises, to move ahead with more systemic reform if they want to secure old-age protection over the long run, while improving labor market incentives in the short run. Health care reform is another way to pursue the latter objective. To make up for persisting inefficiencies in the sector, current health insurance contribution rates are set far higher than those prevailing within the EU. Above and beyond the direct benefits it would bring to the sector, more decisive action on health care reform, by creating room for cutting back the payroll tax, could also contribute indirectly to the overarching growth and convergence the CEECs have set out for themselves. Finally, we have highlighted how key expenditure programs could be redirected toward growth objectives. The education sector, for example, could make a better contribution to CEEC integration into the global knowledge economy if countries took advantage of the passing of a "baby bust" wave to downsize the staffing and facilities of basic and vocational education, and to shift resources toward other educational inputs as well as higher or more generic levels of education. Transport networks can also play a key role in making growth a reality. But their impact will be more powerfully felt if investment decisions are guided by economic rather than by other considerations. In practice, for many countries this would imply prioritizing maintenance expenditures and phasing in new motorway construction only as the underlying traffic justifies it. Similarly, the rates of return on environmental programs can vary within a considerable margin. In some cases they can actually be quite high. Estimates presented in this report suggest that clean air (and to a lesser extent clean water) measures would yield the highest economic benefits in relation to costs, but raise questions about the amount of investment into waste management that would be economically viable. These indications are still tentative, however, and the merit of each individual project will need to be ascertained on a case-by-case basis. But the cause of growth and convergence will be better served if those projects showing the highest return receive first priority. To conclude, the general thrust of expenditure strategies put forward in PEPs appears both appropriate and, at first blush, at least theoretically feasible. As was noted at the outset, - 100- An Expenditure Strategyfor Growth and Convergence factors related to the country's political economy and to the institutional framework of public resource management will play a determining role in framing actual policy choices. This report will have served its purpose if it helps illuminate, and perhaps broaden, these choices and help highlight which courses of action might be practicable, judging by the experience of the CEECs themselves. - 101 - Recent World Bank Technical Papers (continued) No 463 Stephen Foster, John Chilton, Marcus Moench, Franklin Cardy, and Manuel Schiffler, Gionuidzwater zin Rural Developmiienit Facing the Challenges of Supply annd Resouirce Suistainability No 465 Csaba Csaki and Zvi Lerman, eds, Strictuiral Change in the Farming Sectors in Central and Eastern Euroope Lessonsfor EU Accession-Second World Bankl FAO Workshop, Julle 27-29, 1999 No 466 Barbara Nunberg, Readyfor Euitope Plibic Adnuiistiation Reforni and Europeanli Union Accession in Central and Eastern Euirope No 467 Quentin T Wodon with contributions from Robert Ayres, Matias Barenstein, Norman [icks, Kihoon Lee, William Maloney, Pla Peeters, Corinne Siaens, and Shliomo Yitzhaki, Poverty and Policy in Latin America and the Caribbean No 469 Laurian Unnevehr and Nancy Hmrschllorn, Food Safety Issuies in the Developilng World No 470 Alberto Valdes, ed, Agricultural Snpport Policies in Transitioni Econoniies No 471 Brian Pinto, Vladimir Diebentsov, and Alexander Morozov, Dismantling Ruissia's Noiipaynitenits Systeiri Creating Conditionsfor Growth No 472 lit B S Gill, A Diagnostic FranmezWorkfor Reveniue Administrationi No 473 Esen Ulgenerk and Leila Zlaouil, From Transitio,i to Accession Developing Stable aiid Conipetitive Financial Markets in Bulgaria No 474 loannis N Kessides, ed , Hulngary A Regulatory aiid Striuctiural Review0 of Selected Infrastructiure Sectors No 475 Csaba Csaki, Zvi Lerman, and Sergey Sotnikov, Farni Sector Restructuring in Belarus Progress and Constraints No 476 Katherine Terrell, Czech Repuiblic Labor Market Report No 48 L Csaba Csaki, John Nash, Achimll Fock, and Holger Kray, Food and Agriculture in Bulgaria The Challenge of Preparingfor EU Accession No 482 Peter Havlik, Trade anrd Cost Coiripetitiveniess in the Czech Republic, Huingary, Poland, and Slovenia No 483 Mojmir Mrak, Communlntial Infrastructure in Slovenia Suirvey of lnvestmtieiit Needs and Policies Ainmed at En1couiraging Private Sector Puirticipatioli No 484 Csaba Csaki and Laura Tuck, Rutral Developnieuit Strategy Eastern Em tope and Central Asia No 488 Nina Bubnova, Governance Impact oni Private Iiivestnuient No 489 Tim Schwarz and David Satola, Telecoiiiniuinucations Legislation in Transitional and Developing Economies No 490 Jesko Hentschel and Radha Seshagirm, The City Poverty Assessnienit A Priner No 491 Damel Mullei-jentsch, The Devlopment of Electricity Markets in the Eu, o-Mediterraneani Area No 492 Tuntmvate Voravate, Douglas F Barnes, and V Susan Bogach, Assessing Markets for Renewable Energy in Rutral Areas of Northwesterii Chinia No 496 Jerry Lebo and Dieter Schelling, Desigii anid Appraisal of Ruiral Transport Infrastructure Ensuring Basic Access for Rural Comniinitnues No 497 Julian A Lampietti, Anthony A Kolb, Sumila Gulyam, and Vahram Avenesyan, Utilihy Pricing and Poor LessonusfroniuArnienia No 498 Gillian Perkins and Ruslan Yemtsov, Armeenia Restructuring to Suistain Universal General Ediucationi No 499 Rogrigo A Chaves, Susana Sanchez, Saul Schor, and Emil Tesliuc, Fuimancial Mankets, Credit Conistraintts, anid Inivestmtienit 11 Ruiral Roiiiaiiia No 500 Zvi Lerman and Karen Brooks, Tuurkmnenistan An Assessiienit of Leasehold-Based Farni Restruictiring No 501 Aldo Baietti, Private Infrastructture in Enist Asia Lessons Learned in the Aftermath of the Crisis No 505 Ali Hashimn and Bill Allan, T1easuir-y Reference Model --No 506 Omer Gokcekus, Nick Manning, Ranjana Mukherjee, and Raj Nallari, Institutiional Emivirouument amid Public Officials'Performaince in Gu/yania No 507 Ranjana Mukherjee, Omer Gokcekus, Nick Mamning, and Pierre Landell-Mills, Bangladesh The Experience aiid Perceptions of Puiblic Officiiils No 509 World Bank, Kosovo Economtiic and Social Reformsfor Peace and Reconciliation No 510 Anatoly Vinokur, Joana Godinlo, Christopher Dye, Nico Nagelkerke, The TB and HIV/AIDS Epidemics in the Russiant Federation No 512 Geremia Palomba, Milan Vodopivec, Financing, Efficiency, and Equity in Albanian Eduiction No 513 Thomas O'Brien, Christian Filhpov, The Cuirrent Regilatory Franmework Governing Butsiness in Builgaria THE WORLD BANK I818WH Street, N.W Waslhington, D.C. 20433 USA 'l'elephone: 202-473-1000 Facsimile: 202-477-6391 Internet: www.worldbank.org E-mail: feedback@worldbank.org ISBN 0-8213-5368-3